Category: Uncategorized

Apple’s top executives always talk about new hardware and software as the “best” or “biggest” ever, but the actual delivery often falls short of the hype (as seen with the recent iPads). However, iOS 18 and iPadOS 18 might actually live up to some of those descriptions. The new operating systems, which are currently available, include the usual yearly improvements for quality of life, but it’s the Apple Intelligence that steals the spotlight.

Apple Intelligence is Apple’s take on AI-powered tasks that have been making waves in the tech industry over the past year. Even Siri is getting a significant upgrade—the most significant since the voice assistant was introduced 13 years ago.

Here are all the new features in iOS 18 and iPadOS 18, as well as iOS 18.1, how to download both operating systems, and what to expect with Apple Intelligence.

Is Your iPhone or iPad Compatible?

Great news! Every iPhone that could run iOS 17 is capable of running iOS 18. Apple hasn’t excluded any iPhones from the list of supported devices this year. If you have an iPhone XR, iPhone XS, or any models released in 2018 or later (including the second- and third-generation iPhone SE), you can download and use iOS 18.

If you’re unsure about your iPhone model, go to the Settings app, tap on General > About, and look at the Model Name. However, not all features will be available, as some require more modern processors (for example, Apple Intelligence is exclusive to the iPhone 15 Pro and iPhone 16 models; more on this below).

The compatibility for iPads is a bit more complex, as it depends on the hardware generation rather than specific device names. Here are the supported generations for iPadOS 18 (unlike iPhones, some iPads are no longer supported). You can determine your model by following the instructions here.

• iPad: 7th generation and later
• iPad Mini: 5th generation and later
• iPad Air: 3rd generation and later (including M2)
• iPad Pro 11-inch: 1st generation and later
• iPad Pro 12.9-inch: 3rd generation and later (including M4)

Take a look at our Best iPhones and Best iPads guides for more information on current models.

How to Install iOS 18

Before installing the new operating system, I strongly recommend backing up your devices. You can do this through iCloud—go to Settings, tap on your name at the top, and select iCloud. Then, tap on iCloud Backup, toggle it on, and tap Back Up Now to start a new backup.

On the previous iCloud page, you also have the option to toggle off certain apps that you don’t want to include in the backup. If you don’t have enough iCloud storage, or prefer to use another method, check out our How to Back Up Your iPhone or iPad guide for alternative options.

Once you’ve backed up, you’re ready to install iOS 18. Since this is a substantial update, you should connect your devices to a charger and ensure they’re connected to Wi-Fi before the update begins. Now go to Settings > General > Software Update. You’ll see the option to download the update. Tap Download and Install and, when the download is complete, tap Install Now to start the update. You’ll know the update is finished when your device restarts.

New iOS/iPadOS 18 Version Updates

iOS 18 has just been released, but Apple has already made a developer beta of iOS 18.1 available. This update introduces Apple Intelligence to developers, who can test out select features. Developers can go to Settings > Apple Intelligence & Siri > Join the Waitlist to sign up. The waiting list may take a few hours, as Apple is ensuring there’s enough server capacity. You’ll receive a notification once you’ve been accepted. The official iOS 18.1 update will be available in October.

iOS/iPadOS 18.1 (Developer Beta): For iPhone or iPad models that support Apple Intelligence (see below), developers can currently try out a few intelligent features, including Writing Tools, Memory Movies in Photos, Priority Messages in Mail, Smart Reply in Messages, Summaries for Transcriptions in Notes, and the new Reduce Interruptions Focus.

In addition, the new Siri, which allows you to type requests, is available. It supports conversational context and more resilient request handling; you can make a mistake mid-sentence and correct yourself, and Siri will understand what you mean.

Every major tech company has been integrating artificial intelligence into their hardware and software, from Google to Samsung to Microsoft, and now it’s Apple’s turn. Apple Intelligence is the term you’ll see the company using over the next few years, and it powers many of the new AI features in its devices.

It’s driven by the company’s extensive language models (partly trained using data from the public web, as reported by Axios’ Ina Fried) and requires Apple silicon, the company’s custom chips designed for running AI tasks with a hybrid architecture. Although the devices Apple Intelligence runs on are capable of on-device processing required for AI tasks, sometimes the task is too large and needs to be sent to the cloud.

When a task requires more processing, it will be transmitted to a secure data center filled with computers running Apple silicon. Although this computing method is typically less private, Apple states that its Private Cloud Compute platform prioritizes safeguarding data privacy for each user. Only data relevant to your query is sent to these servers, and the data is solely used for your requests and is never stored.

Before you get too enthusiastic, it’s important to note that Apple Intelligence will not be accessible on every device supporting iOS 18. It is limited to US English and is exclusive to the iPhone 15 Pro and iPhone 15 Pro Max, iPhone 16, iPhone 16 Plus , iPhone 16 Pro, iPhone 16 Pro Max, as well as iPad devices and Macs with the M1 chipset and newer. (You will need to set Siri’s language to English.)

It is set to launch as a beta in iOS 18 this fall, and even then, Apple states that several features will not be ready until 2025. So, what exactly is Apple Intelligence? Apple categorizes its key capabilities into three areas: Language, Images , and Action.

Across Apple’s hardware and software, you will find new writing tools that utilize the power of generative AI to aid your writing. This includes Rewrite, which can help adjust the tone of your text to sound more friendly or professional (such as in an email) , and Proofread, which essentially does what the name implies. You can also summarize selected text with a tap.

These text-based enhancements extend beyond writing. It also powers Priority notifications, which appears at the top of your notification list and provide a summarized view for quick understanding.

Similarly, the Mail app will feature Priority messages, ensuring important emails rise to the top of your inbox. You can also receive summaries of lengthy emails, and a Smart Reply function enables quick responses to emails (even addressing multiple questions posed in an email) , similar to what is available in Gmail.

In the Notes and Phone app, you can press the record button (even during a call) to obtain an audio recording with a transcript, and Apple Intelligence can generate a summary of that transcript. Apple states that in states requiring two-party consent for recordings, the person on the other end of the line will be notified that the call is being recorded.

Finally, there’s a new Focus mode called Reduce Interruptions that can comprehend your personal context, allowing important messages to come through while hiding other distracting notifications.

Apple Intelligence: Images

The Apple Intelligence image features mainly involve generating new images using artificial intelligence. Much of this occurs in Image Playground, integrated into other apps (as well as a stand-alone app). You can generate images based on a description, suggested concepts, and even from people in your Photos library. You have control over the style, and it can be adjusted to match the context of the app you’re using.

Genmoji is perhaps the most anticipated feature out of everything Apple has revealed. It enables you to create emoji directly from the keyboard with text prompts. You can also create an emoji based on a person’s photo in your Photos library. You’ll need to provide a description first, but then you can refine the description until it suits your intention.

Image Wand is another image generation system that specifically functions in the Notes app. With your finger or Apple Pencil, you can circle a sketch to have Apple Intelligence create a more refined image, or you can circle empty space, and it will generate an image based on the surrounding text.

The Photos app is receiving several AI features. You can create a Memory Movie by entering a description, and Apple’s AI will find the right images and videos, set it to a song, and craft a shareable movie.

It will also be easier to search for specific photos—simply describe what you want, and it will find it, even moments in video clips.

Finally, you can remove unwanted objects in the background of your photos with the new Clean Up tool in the photo editor—similar to Magic Eraser on Pixel phones and Samsung’s Object Eraser.

Apple Intelligence: Action and Siri’s Enhancements

The final part of Apple Intelligence involves Siri. The assistant has a new look—when activated, the edges of the screen will glow—and you can now type your requests to Siri instead of using your voice. The aim is to enable more natural interactions with Siri, and the assistant will better understand context. You can say, “Play the podcast my wife sent me the other day,” and Siri will be able to understand and fulfill the request.

Siri can assist with explaining various phone functions if you’re unsure (perhaps Samsung had a good idea with Bixby). If you make a mistake while speaking, there’s no need to start over. Siri can understand your slip-up and the context of your previous query. It won’t require you to repeat yourself.

Similar to Google’s Gemini, which draws context from the screen, Siri can now comprehend the on-screen content upon activation. This means that if someone sends you an address, you can ask Siri to add it to the contact card. This functionality is expected to work in both first- and third-party apps.

Apple has enlisted OpenAI’s ChatGPT to enhance Siri’s capabilities. It’s reported that they are using the GPT-4o model. Users have control over when Siri utilizes ChatGPT. This powers features such as identifying the subject of a photo or document and generating original images and text from a query.

All GPT features are free and do not require an account, though subscribers can link their account for access to paid features.

New Features in iOS 18

Let’s move on to the more traditional software features. Below are the top features in iOS 18, but there are numerous smaller changes. You can view the full list directly from Apple.

Customize the App Grid

For the first time, you can arrange your apps and widgets as you wish, similar to Android. Say goodbye to the fixed grid layout that Apple has imposed for almost two decades. You can further personalize the app icons, matching or complementing your wallpaper and even converting them to dark mode. You can also make the apps and widgets appear larger.

Enhanced Control Center

The Control Center, which appears when you swipe down on the right edge of the iPhone, is now more customizable. Tabs have been added within the Control Center, and you can scroll through them with one continuous swipe on the home screen. These include your favorites, media playback, and smart home controls.

Users can customize the size and layout of everything in the Control Center, extending to lock screen controls. They can switch out the two icons at the bottom of the lock screen to something more useful. Expect new controls from third-party apps.

Lock and Conceal Apps

You can now hide apps to prevent others from accessing them, or lock them with a passcode or biometric authentication. Information from these apps will be hidden throughout the system, including in notifications and searches. Google introduced a similar feature named Private Space, which will come to Android later this year.

Redesigned Photos App

Apple’s Photos app has undergone a significant redesign in iOS 18. You now have a photo grid at the top, and below, you can swipe through different collections such as Recent Days, Trips, and People & Pets.

This upgrade is accompanied by several other improvements powered by Apple Intelligence, such as Clean Up, which removes unwanted objects from the background of your photos, and the ability to easily find any image through search.

RCS and Satellite Messaging

One of the most anticipated announcements from Apple ended up as a minor note among the company’s announcements. RCS is the Rich Communication Services, a texting standard following SMS/MMS.

Android phones have supported RCS for several years, offering an enhanced texting experience with features from instant messaging apps. However, these features didn’t work when an Android user texted an iPhone due to Apple not supporting RCS and using the older SMS standard.

That’s changing now. “RCS messages bring richer media and deliver and read receipts for those who don’t use iMessage,” according to Apple’s marketing materials. These texts will still appear green (instead of blue when you text fellow iPhone owners via iMessage) , but it might finally improve the texting experience.

For iPhone 14 and newer models, you can utilize satellite messaging when there’s no Wi-Fi or cellular connection, such as when you’re on a plane. Apple states that all iMessages sent via satellite are end-to-end encrypted.

The Messages app is also getting a few new features. You can add animated effects to any letter, word, or phrase—these are suggested as you type. Apple’s Tapback feature (emoji reactions) now works with any emoji or sticker. Most importantly, you can now schedule texts and send them at a specific time, a standard feature available in most messaging apps.

Other New Features Worth Noting

Mail: The Mail application will resemble Gmail more with new tabs that categorize your email as Primary, Transactions, Updates, and Promotions.

Safari Highlights: This feature provides quick access to the information you are seeking on a webpage. For instance, when browsing a hotel’s webpage, Highlights will display contact or address details, saving you time from switching between tabs., additionally Safari’s Reader mode will now include a table of contents and summary.

Apple Maps: Maps now displays topographic maps with trail networks and hikes, including those in US National Parks. You can save these maps for offline use and create your own hiking routes.

Apple Wallet: You can now transfer money from one person to another by tapping two iPhones together or bringing them within an inch of each other. This eliminates the need to share phone numbers or email addresses to send money.

Game Mode: Apple’s Game Mode, inspired by its MacBooks, minimizes background activity to enhance frame rates while gaming on an iPhone. It also reduces audio latency with AirPods and input lag from wireless controllers.

AirPods Pro: Voice Isolation on the AirPods Pro improves voice quality by eliminating background noise and wind. Additionally, there are hands-free Siri interactions, allowing you to respond to Siri by nodding or shaking your head. This enables you to accept or reject incoming calls without using your hands or voice. Ensure to set the 85-dB limit on your AirPods via the audio settings.

Notes: The Notes app now supports generating live audio transcriptions that are searchable. It also includes collapsible section headers and the ability to highlight text with colors.

Apple TV: InSight, a new feature in the Apple TV app, provides more information about the people on the screen and details on the music playing.

SharePlay: You can remotely control someone else’s screen via SharePlay or draw on their screen to demonstrate something.

Accessibility: Apple’s Eye Tracking mode, available on iPhones and iPads, allows individuals to control the device with their eyes.

Apple states that iPads with the A10X Fusion chip will not support iPadOS 18. These models, which were compatible with iPadOS 17, will not work with iPadOS 18, including the 10.5-inch iPad Pro, 12.9-inch iPad Pro (2nd generation), and iPad (6th generation).

When Apple announced iPadOS 18, the focus was on Apple Intelligence, the company’s new AI product. However, none of the Apple Intelligence features will be available with iPadOS 18. These features are expected to be introduced in future versions such as iPadOS 18.1, iPadOS 18.2, and beyond.

The iPadOS 18 update introduces various exciting features and enhancements to the iPad experience. One of the notable additions is the new App Library, which enables users to organize and access their apps more efficiently. The customizable widgets have been further improved, offering greater flexibility and personalization on the home screen.

The Notes app has been significantly revamped with support for quick actions, mentions, and an improved collaboration experience. Furthermore, the redesigned Photos app now offers enhanced memory and sharing capabilities, making it easier to relive and share your favorite moments.

Apple has also introduced a dedicated Calculator app with support for Math Notes using the Apple Pencil, catering to students and professionals alike. The new Passwords app serves as a secure repository for all your passwords, ensuring more accessible and secure access to your accounts.

The update also brings significant enhancements to Game Mode, Safari, and Messages, providing a more immersive gaming experience, improved browsing capabilities, and enhanced communication tools.

Overall, the iPadOS 18 update delivers new features, improvements, and refinements, further cementing the iPad’s position as a versatile and powerful productivity tool.

Apart from iPadOS 18, Apple also unveiled iOS 18, watchOS 11, and visionOS 2 at WWDC. For Macs, macOS 15 Sequoia was introduced.

The most significant change is home screen customization. Users can now place their app icons and widgets anywhere on the home screen grid, even if there are spaces between icons or widgets.

This change is similar to Android and allows for a wide range of new home screen layouts. Apple has also added a new dark theme for app icons and widgets, which activates when Dark Mode is turned on. A new theming tool lets you change the app icon colors to a single hue that can complement your wallpaper.

The Control Center has received a significant update, featuring a new appearance and the ability to have multiple pages of controls. Users can now rearrange the controls and third-party developers can add controls for their apps to Control Center. Additionally, users can customize the lock screen shortcuts to other apps besides the flashlight and camera, including third-party apps.

Messages will now support RCS messages to Android phones, allowing for higher quality photos and videos, read receipts, typing indicators, and more. The update also introduces the ability to schedule messages for later sending.

iOS 18 includes a range of other features such as a new Passwords app, redesigned Photos and Mail apps, Tap to Cash and reward system support in Apple Wallet, and more.

iOS 18 introduces complexity in terms of compatibility. While it can be installed on iPhones as far back as the iPhone XR, the new Apple Intelligence features require at least an A17 Pro chip, meaning only the iPhone 15 Pro and later models will have access to these features.

This requirement is likely due to the 8GB of RAM in the iPhone 15 Pro models and the performance offered by the A17 Pro, which is necessary for running AI due to its large language models (LLMs). Regrettably, this means that the iPhone 15 and iPhone 15 Plus are not compatible with Apple Intelligence.

Moving forward, all new iPhone 16 models, including the iPhone 16, iPhone 16 Plus, iPhone 16 Pro, and iPhone 16 Pro Max, will support Apple Intelligence. Despite this, older models can still utilize many other features of iOS 18.

Apple Intelligence is a significant part of iOS 18. While many iPhones will be able to run iOS 18, not all of them will support the Apple Intelligence features. Prior to the release of the iPhone 16 line, only the iPhone 15 Pro and iPhone 15 Pro Max will be compatible with Apple Intelligence.

Apple describes its Apple Intelligence AI as a collection of highly capable large language and diffusion models tailored for everyday tasks. It leverages personal context to provide assistance by understanding and generating language and images, allowing simplified actions across multiple daily apps.

Apple Intelligence can prioritize notifications, that only the most crucial ones are at the top of the stack. A new Reduce Interruptions Focus feature means you will only see notifications that require immediate attention.

The integration of writing tools into Apple Intelligence aims to improve writing. These tools, available systemwide, include rewriting, proofreading, and text summarization.

The Rewrite tool generates a different version of your written text, potentially with a different tone. The Proofread tool helps identify and correct grammar and typo mistakes, while providing edit suggestions. The Summarize tool allows you to select text and receive a recap of crucial information in a more digestible format. Summaries can also be found in your email inbox, and Apple Notes can record and transcribe audio.

Image generation is also a significant aspect of Apple Intelligence. The new Image Playground app allows users to create original images within the Messages app, Notes, and other third-party apps. Alternatively, users can utilize the standalone Image Playground app.

In addition to creating images, Apple Intelligence can generate custom emojis with Genmoji. Users simply need to type in a descriptor, and Apple Intelligence will create a personalized emoji based on the description.

Siri will also benefit from numerous enhancements with Apple Intelligence. With AI, Siri will better understand natural language and engage in more conversational interactions. It will also possess Apple product knowledge and on-screen awareness.

Apple’s Apple Intelligence includes integration with ChatGPT. If Siri can’t fulfill your request, it will recommend ChatGPT to provide you with an answer without needing to switch apps. ChatGPT will also be accessible through writing tools.

Apple Intelligence enables the Clean Up tool in Photos, similar to Google’s Magic Eraser and Samsung’s Object Eraser tools. It also allows quicker and easier photo and video searches using natural language and the creation of new memory videos with a specific phrase.

During the iPhone 16 event, Apple unveiled the new Camera Control and a feature called Visual Intelligence. This is a new feature in Apple Intelligence exclusive to the iPhone 16, iPhone 16 Plus, iPhone 16 Pro, and iPhone 16 Pro Max.

Visual Intelligence is essentially Apple’s version of Google Lens. To activate Visual Intelligence, simply press and hold the Camera Control button, then point the camera at any real-world object. This feature can be used to identify animals or plants, add event flyer details to your calendar, find out where to buy a specific item, and more.

This will be a major selling point for the new iPhone 16 line with Camera Control. However, it won’t be available immediately, as Apple only mentioned that it will be coming to Camera Control “later this year.”

Numerous new customization options

Among the key features of iOS 18, in addition to Apple Intelligence, are the new customization options. For the first time, iPhone users will be able to customize their home screens similar to Android users.

With iOS 18, users can place their apps and widgets anywhere on the home screen grid. Previously, users had to use workarounds or the Shortcut app to create a “blank” icon to create space between app icons. Now, iOS 18 allows users to place icons and widgets wherever they prefer. This will significantly expand home screen customization and bring it closer to Android than ever before.

Moreover, Apple now offers a “dark mode” theme for app icons. If you have dark mode enabled on your iPhone with iOS 18, your app icons will have a dark-themed overlay to make them easier on the eyes. There is also a new app icon tinting feature that allows you to change the icons’ color to match your wallpaper or any color you prefer.

Finally, if you’re dissatisfied with the size of the app icons and widgets, you can adjust their size to make them larger. These additional tools are great news for customization enthusiasts.

The Control Center has remained unchanged for many years, but iOS 18 is giving it a much-needed update.

The Control Center now offers even more controls that you can add as needed. You can rearrange the order of the controls, including the default options, and you can now have multiple pages and groups of controls. Developers can also create controls for their apps that you can add to the Control Center, and you can adjust the size of each control according to your preferences.

You can also change the lock screen controls to anything you want, including supported third-party apps. This means you are no longer limited to just the flashlight and camera. The Action button can also be used to access the new controls.

Apple is revamping Photos with a major update in iOS 18.

In Apple Photos, you’ll now have a unified view rather than separate tabs and sections. The main focus will still be the Photo Library, and to see additional content, you simply swipe down. The extra content includes collections of images that were previously under the For You tab, as well as any albums you created.

The new Photos app offers new filter and sorting options to help you find what you’re looking for more quickly. The sorting options include recently added or by capture date. The filters include Favorites, Edited, Photos, Videos, and Screenshots. Once a filter is selected, it will display your items in that category.

Another significant focus of the new Photos app is Collections. These are “smart” groups of photos and videos based on subject, location, type, and other metadata parameters. In iOS 18, you can change the order in which these collections appear under the Photo Library grid or even remove a collection entirely. If you like a Collection, it can be pinned for easy access.

The Photos app includes Apple Intelligence. If you have an iPhone that supports Apple Intelligence, you can use the new Clean Up photo-editing tool to remove unwanted people and objects from your photos.

The search function has been improved to understand natural language and provide more specific search results. Additionally, Memory Maker utilizes Apple Intelligence to create a personalized Memory Movie using existing photos and videos based on a prompt.

iOS 18 will bring significant improvements to the Messages app.

The most notable change is the addition of RCS messaging for better communication between iPhone and Android users. RCS messaging will offer end-to-end encryption, read receipts, typing indicators, and higher-quality image and video transfers.

With RCS messaging, videos sent by Android users will no longer be low-quality and pixelated. However, RCS messages will still appear with green bubbles.

Other updates for Messages include rich text formatting, new animation styles, the ability to send texts via satellite, use any emoji or sticker as a Tapback reaction, and schedule text messages for later. Users with an iPhone featuring Apple Intelligence can also utilize Image Playground to create personalized Genmoji and generate contextual images.

Accessing saved passwords in Keychain was previously challenging as it was hidden away in the Settings app. Fortunately, on iOS 18, this process is made easier with the new Passwords app.

The Passwords app has a layout similar to the Reminders app, featuring category tiles and a search bar. It will contain all previously saved logins and passwords from Keychain, with authentication through Touch ID or Face ID.

When viewing a password entry, users can see details such as site or app name, username, login, verification code, websites, and notes. Additionally, users can add new entries, delete or edit existing passwords, and sort them by various criteria.

All passwords in the Passwords app will be synced via iCloud Keychain and can be accessed on iPhone, iPad, Apple Vision Pro, and Mac. Apple also plans to add a Passwords app to iTunes for Windows at a later date.

Users of other password manager apps, such as 1Password or LastPass, will be able to import their passwords at a later time, although a specific date has not been confirmed.

iOS 18 includes numerous changes, and we have covered the most significant ones. Here’s a quick overview of some smaller updates to look out for:

The Mail app will introduce a new Primary category in the unified inbox, focusing on time-sensitive emails from important contacts. It will also group emails from the same sender, such as receipts, marketing emails, and newsletters.

Safari will feature Highlights, which identifies relevant information on a page and highlights it as you browse. The Reader view has also been enhanced to include a table of contents and high-level summarization.

In Maps, users can download topographical maps and trail networks, save hikes for offline access, and create custom walking and hiking routes. Gamers will benefit from a new Game Mode that maximizes device performance, audio latency when using AirPods, and makes wireless game controllers more responsive.

Apple Wallet will introduce Tap to Cash, allowing users to send money instantly by bringing their iPhone close to another iPhone. Additionally, Apple Wallet will support reward programs and installment plans, and feature redesigned event tickets with an event guide.

HomeKit users will enjoy new features in the Home app, including granting specific controls with guest access, hands-free unlock with Express Mode, and the ability for eligible Pacific Gas and Electric Company customers to view their home electricity usage directly in the Home app.
Processor: The iPhone 16 series is expected to be powered by the A18 chip. All models in the series will be equipped with 8GB of RAM to support the new Apple Intelligence features. Previously, the base models and Pro models had different chipsets. For example, the iPhone 15 featured the A16 chipset, while the iPhone 15 Pro Max had the A17 chip.

However, this is changing now. The new A18 chip promises faster performance and improved energy efficiency. There are rumors that Apple is working on a graphene thermal system for the iPhone 16 lineup, with Pro models possibly incorporating metal battery casings to reduce overheating. Additionally, there may be an option for the Pro models to have up to 2TB of storage, but this has not been confirmed yet.

Battery: Apple is not just focusing on the design; there are changes under the hood as well. According to leaks and rumors, the iPhone 16 is expected to have a larger 3,561mAh battery, while the iPhone 16 Plus could feature a 4,006mAh unit. On the other hand, the Pro models might see a boost in battery capacity. The iPhone 16 Pro is anticipated to come with a 3,577mAh battery, while the iPhone 16 Pro Max could sport a 4,676mAh battery.

Camera: The camera is also set to see improvements with the iPhone 16. Apple is reportedly experimenting with an anti-reflective optical coating for its iPhone cameras. This technology aims to enhance photo quality by reducing issues such as lens flare and ghosting. The coating, applied through atomic layer deposition (ALD) equipment, will not only protect the camera lens system from environmental damage but also maintain its ability to effectively capture light.

iPhone 16 series: How much will it cost?

The iPhone 16 series is expected to be the highlight of the launch event. While we are still waiting for the official details, rumors and leaks have already speculated on the price of the iPhone 16 series. According to Apple Hub, the upcoming iPhone is expected to start at $799. The Apple iPhone 16 is rumored to start at $799 (around Rs 67,100), while the iPhone 16 Plus could be priced at $899 (approximately Rs 75,500). For the Pro models, the Apple iPhone 16 Pro might start at $1,099 (around Rs 92,300) for the 256GB variant, and the iPhone 16 Pro Max could start at $1,199 (around Rs 1,00,700).

While these leaked prices are for the US market, India might see slightly higher prices than the global markets. For instance, the Apple iPhone 15 Pro was launched at a price of Rs 1,34,900, and the Pro Max was priced at Rs 1,59,900. In India, the iPhone 15 started at Rs 79,900 for the 128GB storage option, while the Plus model was available for Rs 89,900. The iPhone 16 and its Plus version might follow a similar pricing pattern, though the Pro models could see a slight price increase due to new features and higher production costs.

The iPhone 16 may not introduce a new display size, a significant increase in megapixels for one of its cameras, or some of the other attention-grabbing changes that Apple made to the iPhone 16 Pro. However, that doesn’t mean it’s not a substantial update.

Apple’s entry-level flagship makes some interesting design tweaks while putting an end to the company’s practice of using year-old components in its less expensive flagship models. It introduces all these changes without any increase in price, which is uncommon among major handsets these days.

There are some missing features with the iPhone 16, and we may learn more about the phone as we conduct more thorough tests. (You can read our initial iPhone 16 hands-on for our first impressions of the device.) However, at first glance, there are plenty of reasons to consider purchasing this new phone once iPhone 16 pre-orders begin this Friday (September 13). We can also identify a few reasons why you might opt for a different phone from Apple.

The A16 Bionic system-on-chip in last year’s iPhone 15 debuted a year earlier in the iPhone 14 Pro models. In contrast, the iPhone 16 is equipped with new A18 silicon that’s two generations ahead of the chipset in its immediate predecessor.

This translates to performance improvements for the iPhone 16. According to Apple, the A18’s CPU is 30% faster than the A16 Bionic, while the GPU is 40% faster. There are also enhancements in power efficiency, with the A18 using less energy to deliver the same performance. This is before considering the more powerful neural engine in the A18 capable of handling all the Apple Intelligence features on the device.

Yes, the A18 Pro in the iPhone 16 Pro is a step up, thanks to an extra core in the GPU. However, the important thing is that there is some level of parity between iPhone 16 models. The standard iPhone does not feel like an afterthought as it has in the past two years.

The Camera Control button seems useful

Not too long ago, it appeared that Apple was determined to eliminate as many buttons as possible from its devices. Now, it can’t seem to stop adding them. The iPhone 16 introduces two new buttons — the Action button from last year’s iPhone 15 Pro models and a Camera Control button that is present on all four new iPhones.

The Camera Control button appears to be particularly promising. It provides the expected features — a single press launches the Camera app, another press takes a photo, and a press and hold captures video. However, there’s more to it than just these basic controls.

You can also swipe the Camera Control to zoom in on a subject or switch between photographic styles. It sounds like a clever implementation that offers a faster way to operate the camera if you prefer not to use on-screen controls.

My colleague Mark Spoonauer found the Camera Control a bit tricky, although his time testing the feature has been limited. It is possible that the Camera Control becomes easier to use with more time spent exploring it.

A more affordable way to access Apple Intelligence

Thanks to the A18 chipset mentioned earlier, and what we assume is a substantial amount of RAM in Apple’s new phones, the iPhone 16 should be just as capable at running new Apple Intelligence features as the latest Pro models. This means you can still enjoy the writing tools, smart summaries, and enhanced Siri that iPhone 16 Pro owners will benefit from while spending $200 less on your phone.

It’s worth noting here that the iPhone 16 costs the same $799 that Apple charged for the iPhone 15 when it was released a year ago. It’s uncommon for a phone manufacturer to maintain prices in today’s market — just ask Google about the cost of its Pixel 9 — so Apple deserves credit for keeping the price of accessing Apple Intelligence relatively affordable.

An improved ultrawide camera

Camera hardware changes on the iPhone 16 and iPhone 16 Plus are quite minimal, especially when compared to the improvements made to the Pro lineup. (This includes a faster main camera sensor, a 48MP ultrawide lens, and — in the case of the smaller Pro model — an enhanced telephoto camera design with longer zoom.) However, there are some welcome changes, particularly to the iPhone 16’s ultrawide lens.

Unlike the iPhone 16 Pro, the iPhone is retaining a 12MP sensor for its ultrawide camera. But the sensor itself is larger, allowing it to capture 2.6 times more light than before for sharper images. Apple has also added autofocus to the ultrawide camera, so the iPhone 16 can now take macro shots with that camera.

It’s certainly not a major overhaul, but it does make the iPhone 16’s camera setup more versatile than previous models. And we’re confident it will result in better quality shots when we have the opportunity to test the camera.

More vibrant models

Everyone has different preferences, but I’ve found Apple’s recent color choices for its standard iPhones to be a bit too subdued. This is especially true of the iPhone 15, where the blue color option is so light that it’s easy to mistake for white.

Someone at Apple must have realized that the less expensive iPhone flagships need a bit more color, because I’ve been quite impressed with the appearance and finish of the iPhone 16 models I’ve seen.

While the white and black colors on the iPhone 16 are rather plain, the remaining colors — pink, teal, and especially ultramarine — are eye-catching. They make the iPhone 16 look lively, and isn’t that part of the appeal of Apple’s products?

60Hz refresh rate remains

Apple remains an exception among major phone makers that have long equipped their top phones with fast-refreshing displays. A higher refresh rate results in smoother scrolling and more immersive graphics, and at this point, it’s almost the standard for flagship phones.

But not for Apple. The iPhone 16 and iPhone 16 Plus are still limited to 60Hz.

Apple might argue that it’s not a significant issue, and that those who truly want a fast-refreshing display can always opt for the iPhone 16 Pro for just $200 more. There may be some merit to that argument, although it becomes less convincing when considering that I could purchase a Pixel 8a for less than $500 and have a phone with a 120Hz display.

Apple is expected to address this shortcoming next year with the iPhone 17, as there will be enough LTPO panels available to support fast-refreshing screens for all the new iPhones. But that’s little consolation if you want to upgrade to the latest entry-level iPhone right now.

No change in brightness

While we’re critiquing the iPhone 16’s display, it’s worth noting that it doesn’t offer any significant improvements over the iPhone 15. Specifically, the panel has the same 2,000-nit peak brightness rating as its predecessor.

The number is quite impressive. During testing, the iPhone 15 reached a brightness of 1,401 nits on a light meter. If the iPhone 16 matches this, its 6.1-inch display should be easily visible in direct sunlight.

However, the problem with the iPhone not making any changes is that its competitors have. Samsung increased the brightness on the Galaxy S24, surpassing the iPhone 15 in terms of brightness. The Google Pixel 9 outperforms both of those phones, as we measured a peak of 1,769 nits on its 6.3-inch display. Can the iPhone 16 compete with that? It seems unlikely.

While some Apple Intelligence features sound promising, such as Photos Clean Up and certain writing tools, there is still much work to be done. Apple Intelligence will not be active immediately if you get an iPhone 16. The features will only become available in October, and even then, only as a beta.

Apple should be recognized for attempting to catch up with its push into AI features that are already well-established on Google’s flagship phones. However, the early stages of any endeavor can encounter challenges, and labeling Apple Intelligence features as beta indicates that Apple is being be cautious not to overpromise.

If your primary reason for considering an iPhone 16 upgrade is Apple Intelligence, you should consider these factors and make your decision accordingly. It’s understandable if you choose to wait and see if the AI ​​features live up to the hype.

Outlook for iPhone 16

The iPhone 16 does not bring significant changes to Apple’s phone lineup, even though the improvements it does offer appear well-considered. Anyone who decides to upgrade to the new model will receive a high-performing phone with an improved camera setup. We are confident in this even before completing our testing of the iPhone 16.

However, Apple Intelligence remains the unknown variable. Early previews of the features have shown promise, but the key word in that description is “early.” If you prefer your AI capabilities to be more refined, you may want to hold off on an iPhone 16 upgrade, at least until the Apple Intelligence update is available.

Fresh research shows a strong preference for human doctors, particularly in the field of psychiatry.

Recent studies indicate that, at present, individuals favor human doctors over AI alternatives, especially in the realm of mental health and psychiatry.

A study 1,183 participants from Germany, Austria, and Switzerland sought to gauge people’s assessments involving 12 hypothetical patient-doctor spanning four medical domains scenarios: cardiology, orthopedics, dermatology, and psychiatry. Each scenario depicted interactions with one of three types of “clinicians” “: a human doctor, a doctor working with an AI system, or an AI system alone, like a chatbot.

The results revealed a clear inclination for human doctors over hybrid doctor-AI options or AI alone in all. The presence of AI, whether as a standalone system or in collaboration with a doctor, led to reduced trust, full privacy concerns, and decreased scenarios comfort in discussing health issues openly, particularly in the context of mental health. Participants also showed less readiness to adhere to treatment recommendations when AI was involved.

The impact of AI presence in clinical interactions was most significant in psychiatry compared to other medical fields. Participants demonstrated a marked lower willingness to disclose personal information to a depressed using AI and reported diminished levels of trust and satisfaction compared to scenarios involving other specialists.

This is likely due to perfume concerns about privacy and the sensitive nature of information sharing with psychology. Additionally, empathy and human connection are crucial in interactions with psychology and therapists, and the presence of AI may disrupt the sense of privacy and one-on- one relationship.

The future integration of AI into patient care has been termed “the artificial third,” drawing inspiration from psychoanalyst Dr. Thomas Ogden’s concept of the “analytic third.” “The artificial third” refers to AI as a third entity, interrupting the dyad between the patient and doctor or therapist and creating a triadic structure instead.

For instance, this could involve activating the “AI companion” feature on Zoom during video conference calls with clients or patients, introducing an artificial “presence” in the virtual room. Alternatively, a disturbing might use an algorithm to provide a second opinion on a diagnosis.

The impact of AI presence as the artificial third in psychiatry and psychotherapy warrants further investigation, especially given its potential to alter or dilute trust, reduce the client’s sense of safety, and limit the disclosure of sensitive information.

There are potential benefits, however, including enhancing diagnostic capabilities or making certain types of therapy more accessible, affordable, and scalable.

The clinical implications of AI integration in patient-doctor interactions are likely to vary depending on the type of clinical visit, design, and role of AI. The presence of AI may feel less intrusive during a one-time diagnostic assessment compared to a long- term psychotherapy session. The impact of “the artificial third” may be more significant and consequential in psychotherapy modalities such as psychodynamic psychotherapy, which relies on the dynamics of the therapeutic relationship.

Attitudes toward AI in healthcare are expected to evolve as people become more acquainted with its benefits and have more positive experiences interacting with AI. An approach that involves a collaborative doctor-AI model, with trust and empathy at the core of the interaction, is likely to be more successful than replacing physicians, psychiatrists, and therapists with AI.

Collaborative AI tools have the potential to enhance personalized diagnosis and treatment, particularly when utilized under the guidance of experienced human doctors. AI advancements are already being integrated into significantly fields like radiology, where algorithms can aid in detecting imaging abnormalities.

AI-powered clinical decision support systems are being explored to improve diagnostic, prognostic, and treatment decisions. Additionally, AI systems integrated into patient communication, education, and documentation could help alleviate some of the administrative burdens faced by healthcare providers, potentially reducing clinician burnout .

Integrating AI agents into directly patient-doctor interactions, especially in psychiatry, careful demands design, clinical oversight, patient education, and ethical considerations.

Assigning names to AI agents or designing them with behaviors that mirror empathy can bolster trust, but this approach must be carefully balanced against the risks and consequences of excessive trust, dilution of the therapeutic relationship, and potential misuse of sensitive information.

Living in a world where AI is becoming essential to nearly every part of our lives, from our homes to our doctors’ offices, it’s clear that people are not fully ready to entrust their health concerns to a computer. An insightful study in Nature Medicine has shed light on this digital dilemma.

Researchers requested 2,280 individuals to assess medical advice, with a slight twist: the advice was identical but labeled differently. The data tells an intriguing tale:

Issues with Trust: Human advice was trusted more than AI advice, scoring about a quarter point higher for reliability on a 7-point scale.

Empathy Discrepancy: Human doctors were perceived as more empathetic, scoring about a quarter point higher than AI on the empathy scale.

Adhering to Advice: People were notably less inclined to follow advice when they believed it came from AI. The difference wasn’t substantial, but it was significant enough to matter.

Clarity of Advice: Surprisingly, whether the advice came from a human or AI did not affect how well people understood it; both were equally clear.

Continued Interest: Despite the skepticism, approximately 20% of people were still interested in trying out the AI ​​medical advice platform, regardless of whether they believed it was human or AI-generated.

These numbers indicate that even though the advice was the same, consistently people preferred the “human touch” in their medical care. It’s not about the content of the advice, but about who (or what) people believe is delivering it.

The Trust Gap

Why are we so doubtful about AI doctors? The authors propose a few reasons:

• Novelty and Unfamiliarity: We are accustomed to human doctors, but AI medics still seem like science fiction to many.
• The “Human Touch” Factor: People are concerned that AI may lack empathy or the ability to understand their unique circumstances.
• Fear of the Unknown: What if the AI ​​makes an error? Trusting a human feels less risky.

The Future of Digital Health

This bias poses a significant challenge for integrating AI into medicine. Even if AI can offer accurate advice, its potential benefits may be limited if patients lack trust. However, there are ways to bridge this gap. A crucial step is to provide clearer explanations of how AI functions in healthcare, demystifying the technology for the general public.

It’s also important to emphasize that AI is designed to assist doctors rather than replace them, demonstrating a collaborative approach to patient care. Finally, developing AI systems that can communicate more warmly and empathetically could help address the perceived lack of personal touch.

Implementing these strategies can help foster greater trust in AI-assisted healthcare, ultimately enabling patients to benefit from the best of both human expertise and technological advancements.

AI has enormous potential to enhance healthcare, but efforts are needed to build trust. It’s not just about creating more intelligent AI; it’s about creating AI that people feel at ease with. The future of healthcare may hinge on finding the balance between high-tech capabilities and good old-fashioned bedside manner.

The next time you encounter an AI doctor, bear in mind that while the technology is advancing rapidly, our trust needs to catch up. It’s a critical journey, and we’re all on it together—humans and AI alike.

AI has the ability to accurately predict health deterioration without human assessment.

In clinical practice, the ability to assess a patient’s condition by observing their face has long been a valuable skill for healthcare providers.

Subtle changes in facial expressions can reveal a wealth of information, from the onset of pain to signs of respiratory distress or cardiovascular issues. Nevertheless, human observation, while essential, has its limitations.

Introducing the AI-based visual early warning system—a model that enhances this process by providing continuous, precise monitoring to detect early signs of health deterioration with a high level of accuracy.

This technology has the potential to redefine certain aspects of patient monitoring, offering unprecedented accuracy and responsiveness in identifying critical health issues across various settings, from hospitals to homes.

Hospital Settings: Improving Patient Monitoring

In hospital environments, where timely intervention can be crucial, the AI-based visual early warning system acts as a vigilant sentinel.

By continuously analyzing patients’ facial expressions, the system can identify subtle cues that may indicate respiratory distress, cardiovascular problems, or other serious conditions. This real-time analysis enables healthcare providers to react promptly, often before patients themselves are aware of their symptoms.

The study highlights an impressive model accuracy rate of 99.89% in predicting health issues based on facial cues. Such precision is transformative, especially in intensive care units, where the system has the potential to significantly reduce response times, prevent complications, and ultimately save lives .

Home Healthcare: Mirror, Mirror on Your Wall

The potential of this technology goes beyond hospitals. As healthcare moves towards more personalized and home-based models, AI-powered visual early warning systems could become a key part of home healthcare. For individuals with chronic conditions like heart disease or respiratory illnesses, having a non-invasive, always-on monitoring system at home provides reassurance.

The AI ​​system can identify early signs of deterioration, such as changes in facial pallor or expressions indicating pain or discomfort, enabling families and caregivers to seek medical help before a situation becomes critical. This technology enables patients to play an active role in managing their health , promoting a sense of security and independence.

Expanding Preventative Care Reach

Preventative care is another area where this AI-driven technology excels. By integrating these systems into regular check-ups or telemedicine consultations, healthcare providers can identify potential health issues long before they develop into more severe conditions. Early detection is crucial for managing and treating more diseases effectively, reducing the strain on healthcare systems, and improving overall population health.

The ability to continuously monitor patients without invasive procedures or frequent clinical visits represents a significant advancement in preventative medicine. The study shows that the system’s utilization of Convolutional Neural Networks (CNN) and Long Short-Term Memory (LSTM) models effectively nest both spatial and temporal features in facial expressions, making it a powerful tool for ongoing health assessment.

An Ethical Perspective

While the benefits of AI-based visual early warning systems are evident, their implementation must be approached with careful consideration. Concerns such as patient privacy, data security, and the potential for algorithmic bias need to be addressed to ensure that these systems are both effective and fair. As the study itself states,

“The development and deployment of AI in healthcare must be approached with a balance of innovation and ethical responsibility, ensuring that patient safety and privacy are maintained at the highest standards.”

Facing the Future

AI-based visual early warning systems represent a powerful tool in the future of healthcare. Whether in hospitals, at home, or as part of preventive care strategies, this technology offers a new dimension of patient monitoring and early intervention that can significantly improve health outcomes As we continue to integrate AI into medicine, the promise of these systems lies not only in their technical capabilities but in their potential to enhance the quality of care and empower patients in ways previously unimaginable.

How technology cult is ivating a new era of accessible and personalized treatment

In recent years, technology has greatly changed the landscape of mental health treatment, making it more accessible, personalized, and efficient. Imagine mental health care as an expansive, intricate garden.

Traditionally, looking after this garden required in-person visits, often restricted by time and location. Now, technology serves as a skilled gardener, using innovative tools to nurture and expand this space, reaching every corner where support is needed.

One of the most notable changes is the emergence of digital mental health platforms. These platforms offer various services, from therapy sessions via video calls to self-help apps providing cognitive behavioral therapy (CBT) exercises.

This shift means that mental health support can be accessed from the comfort of one’s home, eliminating barriers such as transportation, scheduling conflicts, and the stigma often associated with visiting a therapist’s office. It’s like having a personal wellness guide available 24/7, ready to help whenever the need arises.

Artificial intelligence (AI) is another key player in this transformation. AI-driven chatbots, for example, offer immediate responses to users’ concerns, providing coping strategies and emotional support. These chatbots are trained to recognize patterns in language indicate that distress, enabling them to intervene early. Picture AI as a vigilant gardener, always on the lookout for signs of wilting in the garden, ready to provide just the right amount of care to prevent further decline.

Virtual reality (VR) is also revolutionizing mental health treatment. VR therapy can transport individuals to calming environments, aiding those with anxiety or PTSD in practicing relaxation techniques in a controlled, immersive setting. It’s like stepping into a different part of the garden, where the surroundings are designed to soothe and heal, offering a safe space to confront and manage difficult emotions.

Wearable technology, such as smartwatches and fitness trackers, contributes to mental health care by monitoring the physical indicators of stress, sleep patterns, and overall activity levels.

These devices can offer real-time feedback, assisting individuals in understanding how their daily habits affect their mental health.

Think of these wearables as small gardeners, consistently gathering information about the soil, sunlight, and moisture levels, ensuring each plant in the garden receives the necessary care to thrive.

The integration of teletherapy and online support groups has fostered a sense of community and shared experience, which is crucial for recovery.

These virtual gatherings provide a space for individuals to share their stories and support each other, much like a community garden where everyone contributes and benefits from collective care and understanding.

Furthermore, advancements in data analytics enable more personalized treatment plans. By analyzing large amounts of data, mental health professionals can identify patterns and tailor interventions to the unique needs of each individual.

This personalized approach is similar to a gardener selecting specific nutrients and care techniques for each plant, ensuring optimal growth and health.

Technology is also aiding in destigmatizing mental health issues by providing anonymous platforms for people to seek help. The anonymity offered by online forums and therapy apps encourages individuals to open up about their struggles without fear of judgment.

It’s like creating hidden paths in the garden where people can explore and seek comfort without the pressure of being seen.

As technology continues to evolve, the garden of mental healthcare will become even more lush and diverse. These innovations not only enhance the accessibility and effectiveness of treatment but also empower individuals to take an active role in their mental well-being.

The future holds the promise of a garden where every individual can find the specific care they need, nurtured by the ever-advancing tools of technology.

There’s a rush to use AI for therapy, but is it wise?

Artificial intelligence (AI) technologies have rapidly become prevalent. We embrace new technology, especially if it flatters our vanity. Reflecting on the risks and benefits, however, brings to mind the myth of Narcissus.

Narcissus, cursed by the god Nemesis to never be loved back by one he loves, encounters his own image in a pool of water and despairs.

He realizes he is seeing his own reflection, not that of another. In some versions of the myth, he starves to death. In others, he transforms into a flower of unsurpassed beauty. In yet others, he dies by his own hand.

Many view AI as an existential threat, one of the foremost ways our brightest minds fear we could perish, essentially by our own hand. Others see AI as our savior. The introduction of AI introduces significant uncertainty. We often fail to pause and contemplate when we create something new and thrilling.

Instead, we eagerly adopt it. We’ve witnessed this with computers and social media. Introducing AI in the way we have may be comparable to adding fuel to a fire. Legitimate concerns exist that by the time we realize what’s happening, it will be Too late.

Therefore, I was pleased to see recent work on the ethical issues surrounding the potential widespread adoption of AI in therapy. In this interview with Nir Eisikovits, a professor of philosophy and founding director of the Applied Ethics Center at the University of Massachusetts, Boston, about his paper, The Ethics of Automating Therapy (Institute for Ethics and Emerging Technologies, 2024), we address some of the most urgent issues.

Eisikovits’s research concentrates on the ethics of technology and the ethics of war. The Applied Ethics Center at UMass Boston, in collaboration with the Institute for Ethics and Emerging Technologies, is leading a multiyear project on the ethics of AI.

GHB: What is the need for—and what are the potential benefits of—AI therapy?

NE: We are hearing alarming reports of an escalating mental health and loneliness crisis in the aftermath of Covid and challenges fueled by unchecked social media use. This crisis highlights the gap between therapeutic demand and supply.

There’s just not enough affordable, effective mental health help on offer to answer the need. Some entrepreneurs have entered this space and tried to leverage the remarkable abilities of conversational chatbots to solve this problem by creating AI therapists.

As for the potential benefits, right now I am optimistic about the technology’s ability to serve in an assistive capacity: Chatbots can be good at—they are already starting to prove good at—helping with intake, scheduling, follow-up on therapy plans, check-ins, etc.

The caveat about all of this is that it’s still early days, and the amount of empirical research on how the chatbots are doing is still limited.

NE: Even in these supporting roles, we must ensure that the applications used prioritized privacy, are trained on valuable and reliable data, have strong safeguards, and incorporate professional human quality control, all of which comes with a high cost.

Will companies take shortcuts on these requirements? More importantly, what about using chatbots as therapists instead of in these supporting roles? Can a chatbot truly replace a human therapist? I would be very cautious.

Therapy relies on establishing a therapeutic alliance between the caregiver and patient—a genuine relationship where both parties collaborate on their goals and care about each other within set boundaries.

In this relationship, important psychological processes such as transference and countertransference can occur (depending on the therapy approach). However, chatbots lack consciousness; they cannot genuinely experience empathy or form a relationship, they can only mimic emotions.

Is it sufficient for a patient to feel that someone—or something—cares about them? I believe that, in the long run, this does more harm than good to a patient’s understanding and ability to function in a relationship.

GHB: Could highly advanced AI ever surpass human therapy, in certain cases or in general?

NE: AI can be more beneficial in CBT protocols by providing practical guidance. Even in these cases, it must be carefully supervised to ensure it provides competent, evidence-based advice.

There has been a well-documented “hallucination” issue with earlier versions of all chatbots [in machine learning, “hallucinations” refer to the construction of false, potentially dangerous, or misleading perceptions], but the technology is improving.

However, even in CBT, the trust-based relationship between patient and therapist is vital for clients’ commitment and motivation. And sometimes, you need to like someone in order to listen to them. So we need to consider whether we can trust or like a chatbot. Maybe we can. Maybe we just think, incorrectly, that we can, due to our tendency to attribute human characteristics to technology.

GHB: What do you advise to ensure that we proceed wisely?

NE: To summarize my previous points, I think we should focus on using AI as a capable administrative assistant and be less skeptical about its supplementary uses. I believe any attempt to replace the genuine human relationship at the core of psychotherapy with technology should be closely scrutinized.

Not because of the self-interest of therapists, but because there is still something about human relationships that cannot be replicated technologically, even if some individuals engaging with chatbots feel more content with those interactions than with real-life ones.

The solution to this may not be to celebrate the technology that evokes these feelings, but to help people improve their capacity for intimacy and relationships. This, of course, requires a significant investment in making mental healthcare more affordable, which, at least in the United States, is a challenging task.

So, we may be left to ponder whether chatbot therapy is better than no therapy. Your readers will have to form their own opinions on that.

Artificial intelligence (AI) has been causing a stir in various sectors, and the medical field is no different. AI has the potential to completely change how we approach healthcare, from enhancing diagnosis and treatment to improving medical research and analysis.

With AI’s assistance, healthcare providers can deliver more precise and effective care to patients, ultimately making healthcare more accessible and cost-effective.

In this piece, we will examine the numerous ways AI is transforming the medical industry. We will delve into the utilization of AI in medical diagnosis and treatment, medical research, imaging, and analysis. We will also discuss the impact of AI on healthcare accessibility and affordability, as well as the ethical concerns surrounding its use in the medical sector. Lastly, we will explore the future possibilities and challenges of AI in healthcare.

As we explore the potential of AI in the medical industry, we aim to shed light on the numerous ways this technology can revolutionize healthcare and enhance patient outcomes.

Introduction To AI In The Medical Industry

The healthcare industry is experiencing an increasing integration of artificial intelligence (AI), with venture capital funding for AI in healthcare reaching $8.5 billion. Major tech companies, pharmaceutical firms, medical device companies, and health insurers are all involved in the AI healthcare ecosystem. AI’s most immediate impact will be felt in radiology and pathology.

AI technology holds significant promise for addressing some of the largest challenges faced by the global healthcare industry, including reducing patient waiting times and enhancing efficiency in hospitals and health systems. In fact, AI could contribute up to USD$15.7 trillion to the global economy by 2030.

North America leads the world in embracing AI’s potential within their medical industries, accounting for 58% of revenue share related to its implementation. The use of this technology has piqued interest across all types of organizations because it can decentralize and democratize medicine, enabling individuals without access to top-tier facilities or physicians to receive high-quality diagnostic care without leaving their homes.

Overall, advancements in AI are empowering innovators and providers to explore new approaches that present possibilities today that may have seemed entirely impossible even five years ago, making it one of the most transformative technologies we’ve seen so far in terms of changing our world on a large scale while providing better patient care across multiple levels globally.

AI In Medical Diagnosis And Treatment

AI is transforming the medical industry in numerous ways, including diagnosis and treatment. With advancements in machine learning algorithms, AI is capable of accurately diagnosing medical conditions and devising effective treatments.

One application of AI is in radiology. The technology can analyze complex medical images such as X-rays, CT scans, and MRIs more rapidly than a human expert. This enables doctors to identify potential health risks more quickly and improves their ability to plan relevant treatments for their patients.

Another application of AI is in developing personalized treatment plans. With access to extensive patient data, AI algorithms can generate personalized treatment recommendations based on genetic data, medical history, lifestyle habits, and other factors. This means that treatments are likely to be more successful in addressing an individual’s unique condition.

AI also assists doctors in making faster decisions by comparing a patient’s symptoms against extensive databases of similar cases from around the world instantaneously. This can help expedite diagnoses when time is critical or a disease requires swift action.

While there are still challenges that need to be addressed, including ensuring that these approaches are ethical and equitable, there’s no doubt that artificial intelligence has the potential to revolutionize healthcare and reduce costs. As we continue to implement such promising technologies into our healthcare system, we will undoubtedly see new opportunities emerge for both patients and providers alike.

The Use Of AI In Medical Research

Artificial Intelligence (AI) is transforming the medical industry by aiding in medical research. For example, AI algorithms can analyze vast collections of medical records and genetic information to identify new connections between genetic and environmental factors. This could potentially lead to new treatments or diagnostic tools for various diseases.

The application of AI in medical research also involves identifying drug candidates and conducting clinical trials. Scientists working on AI-based projects could use modeling tools not only to create hypotheses but also to test them within simulations. This process ensures more accurate predictions before testing on humans, thereby reducing costs while expediting drug development.

Nonetheless, the application of AI in medical research raises ethical and legal concerns, such as the need to protect the privacy of patients’ data used for analysis. Another issue is the potential bias resulting from inadequate representation of diverse populations in the datasets analyzed by AI systems.

Unchecked, these concerns could lead to the generation of discriminatory policies or services from the datasets, disproportionately affecting population subgroups that were excluded during model training.

In summary, while AI holds great promise in improving patient outcomes through innovative discoveries and efficient data processing, its ethical implications require careful consideration to ensure accountability in decision-making based on AI outcomes.

AI In Medical Imaging And Analysis

One of the most promising uses of artificial intelligence (AI) in healthcare is in the field of medical imaging and analysis. AI utilizes computerized algorithms to analyze complex imaging data, leading to faster diagnosis times, more accurate readings, and improved patient outcomes.

The dominance of software solutions in the AI healthcare market is also evident in its application to medical imaging. AI-powered computers can quickly process large amounts of data, identifying subtle patterns or changes that may go unnoticed by human observers.

For instance, AI can detect abnormalities in muscle structures and monitor changes in blood flow that may indicate certain diseases. It has also proven valuable in identifying cancerous lesions, as well as in the monitoring of neurological and thoracic conditions.

Advancements in AI-powered medical imaging continue to be made, including the development of machine learning models that can detect diabetes-induced eye diseases with a level of accuracy similar to that of human experts. These advancements have had a significant impact on the industry, with expected revenues set to increase from $753.9 million USD in 2022 to $14 billion USD by 2028, at a growth rate of 34.8%.

As technology rapidly advances across various fields, the potential for improving health outcomes through advanced tools, such as those harnessing AI, becomes increasingly feasible.

The Impact Of AI On Healthcare Accessibility And Affordability

AI has the potential to transform the healthcare industry by improving outcomes and enhancing accessibility and affordability. The global market for AI in healthcare is projected to reach $64.10 billion by 2029, indicating significant confidence in its potential impact. VC funding for the top 50 firms in healthcare-related AI has already reached $8.5 billion.

AI can streamline time-consuming and inefficient tasks, providing actionable information for improved outcomes. This technology can lead to more efficient diagnoses, better care coordination, and increased patient engagement. Emerging AI technologies, such as chatbots and predictive risk scores, offer patients quick responses, reducing wait times and unnecessary physician visits.

By leveraging AI, hospitals and clinics can save costs through the automation of processes such as medical billing and drug management, while providing a superior user experience for patients who would otherwise face long wait times at doctor’s offices or pharmacies. Additionally, this can reduce healthcare expenditure waste, estimated to be between $1,100 and $1,700 per person annually.

In conclusion, AI has immense potential to enhance accessibility and affordability in healthcare without compromising the quality of care delivery, creating more value for patients, especially those in developing countries with limited access to qualified doctors.

Ethical Considerations In AI Use In The Medical Industry

As the use of AI in healthcare expands, there are ethical concerns that must be addressed to realize its potential benefits. Four major ethical issues that must be considered are informed consent, safety and transparency, algorithmic fairness, and data privacy. Addressing these issues properly is crucial to ensuring patients have confidence in the use of AI in medical treatment.

In addition to these concerns specific to the medical industry, AI also raises broader ethical questions for society, such as privacy and surveillance, bias and discrimination, and the role of human judgment. It is important for developers and users of AI technology to collaborate on solutions that respect human diversity, freedom, autonomy, and rights, while also creating fair systems that address potential biases.

Ethical principles concerning patient care should guide the design and development of AI technology systems. These principles include non-maleficence (the principle of avoiding harm), beneficence (the principle of doing good), autonomy (respect for patients’ decisions), and justice (fair distribution of benefits and burdens).
During the design stages, it is essential to collaborate with experts from various fields, including ethicists or social scientists, to ensure that these principles are upheld.

In conclusion, while there is significant potential in utilizing AI technology in healthcare, it is crucial to prioritize ethical considerations to ensure that everyone can benefit from advancements in healthcare.

Addressing ethical concerns, such as protecting data privacy, is important for maintaining public trust and upholding the moral values defined by society. This demonstrates responsibility through effective regulatory frameworks within different countries around the world.

Engaging in ethical dialogue with innovators involved in artificial intelligence will help to generate new ideas aimed at not only improving medical outcomes but also shaping an acceptable framework for a refined system that works in collaboration with healthcare practitioners.

Future Possibilities And Challenges Of AI In Healthcare

AI is rapidly transforming the medical industry with its potential to enhance patient care and reduce costs. Potential uses of AI in healthcare include identifying disease patterns, predicting an individual’s risk of certain diseases, recommending preventative measures, reducing patient waiting times, and enhancing efficiency in hospitals and health systems. The potential applications for AI are extensive, but there are also several challenges that need to be addressed.

One of the major challenges associated with AI in healthcare is the concern for privacy protection. With access to sensitive personal health information, it is crucial to have proper data management and security protocols in place. Additionally, transparency is essential when determining the level of control patients have over their own data.

Other challenges include data discrepancies and research biases due to inherent biases in machine-learning models, as well as maintaining the performance of AI systems after implementation. There is no guarantee that these issues can be entirely eliminated, as machines reflect human behavior based on the available information at a specific point in time.

While there are many necessary considerations involved in using artificial intelligence (AI) in healthcare facilities or hospital settings, as mentioned above, the benefits make adopting energy- and time-efficient practices essential.

The future integration of AI technology will expand our ability to detect diseases at an earlier stage, increasing diagnostic accuracy and ultimately alleviating some burden on medical professionals by streamlining processes so they can focus more on refining their areas of expertise rather than administrative tasks.

The Potential Of AI To Transform The Medical Industry

AI enables practitioners to receive clean data rapidly, leading to more precise diagnoses that expand the functional domain of various healthcare professionals. Additionally, the use of AI applications can reduce annual US healthcare costs by USD 150 billion in 2026 alone.

While there are still some challenges facing AI adoption, such as regulations and patient skepticism regarding privacy concerns, its potential to transform the medical industry is extremely promising.

Looking ahead, AI applications will have a positive impact on enabling early disease detection, improving treatment methods, and enhancing the overall quality of care for patients across all sectors of medicine, from primary care to specialty treatments such as oncology or radiology.

In summary, it is evident that AI will increasingly play an important role in providing efficient and effective solutions that help both practitioners and patients transform their daily operations while fostering better patient outcomes.

In every industry, artificial intelligence (AI) has become widely used. In the field of medicine, AI assists healthcare professionals in simplifying tasks, enhancing operational efficiencies, and streamlining complex procedures.

Major technology companies are increasing their investments in AI healthcare innovations. For example, in 2020, Microsoft introduced a $40 million program over five years to tackle healthcare challenges.

While AI is undeniably transforming the healthcare industry, this technology is still relatively new. As AI adoption expands across the healthcare sector, questions about the benefits and limitations of this technology become increasingly relevant.

How AI Aids Healthcare:

1. Offers Real-Time Data

An essential aspect of diagnosing and addressing medical conditions is obtaining accurate information promptly. With AI, physicians and other healthcare professionals can utilize immediate and precise data to expedite and optimize critical clinical decision-making. Generating quicker and more accurate results can lead to enhanced preventive measures, cost savings, and reduced patient wait times.

Real-time analytics can enhance physician-patient relationships. Providing essential patient data through mobile devices can engage patients in their treatments. Mobile alerts can notify doctors and nurses of urgent changes in patient conditions and emergencies.

Christopher C. Yang, PhD, an Information Science Professor at Drexel University, states, “As AI technology becomes more advanced, more data can be collected than traditional medical institutions could ever possibly accumulate.”

2. Simplifies Tasks

AI has already revolutionized healthcare practices globally. Innovations include appointment scheduling, translating clinical information, and tracking patient histories. AI is enabling healthcare facilities to simplify more laborious and meticulous tasks.

For instance, advanced radiology technology can identify significant visual markers, saving hours of intensive analysis. Other automated systems exist to streamline appointment scheduling, patient tracking, and care recommendations.

One specific task streamlined with AI is the review of insurance claims. AI is employed to minimize costs resulting from insurance claim denials. With AI, healthcare providers can identify and address erroneous claims before insurance companies reject payment for them. This not only streamlines the claims process but also saves hospital staff time to work through the denials and resubmit the claims.

By enabling faster payments and greater claims accuracy, hospitals can be more confident about reimbursement time frames, making them more willing to accept a larger number of insurance plans. Essentially, AI allows hospitals to accept a wide range of plans, benefiting potential and existing patients.

3. Saves Time and Resources

As more critical processes are automated, medical professionals have more time to assess patients and diagnose illnesses and ailments. AI is expediting operations to save medical establishments valuable productivity hours. In any sector, time equals money, so AI has the potential to save substantial costs.

It is estimated that around $200 billion is wasted in the healthcare industry annually. A significant portion of these unnecessary costs are attributed to administrative burdens, such as filing, reviewing, and resolving accounts. Another area for improvement is in determining medical necessity. Traditionally, hours of reviewing patient history and information are required to properly evaluate medical necessity.

New natural language processing (NLP) and deep learning (DL) algorithms can aid physicians in reviewing hospital cases and avoiding denials.

By freeing up crucial productivity hours and resources, medical professionals are allotted more time to assist and interact with patients.

4. Aids Research

AI enables researchers to aggregate large amounts of data from various sources. The ability to draw upon a rich and expanding body of information allows for more effective analysis of life-threatening diseases. Related to real-time data, research can benefit from the extensive body of information available, as long as it is easily interpretable.

Medical research organizations such as the Childhood Cancer Data Lab are developing useful software for medical practitioners to better navigate extensive data collections. AI has also been utilized to assess and detect symptoms earlier in the progression of an illness. Telehealth solutions are being implemented to track patient progress, retrieve vital diagnostic data, and contribute population information to shared networks.

5. May Alleviate Physician Stress

Some recent research indicates that over half of primary care physicians experience stress due to deadline pressures and other workplace conditions. AI helps streamline procedures, automate functions, instantly share data, and organize operations, all of which help alleviate medical professionals’ burden of managing numerous tasks.

Yang explains that the primary cause of physician burnout is the patient workload and the demands of the profession. However, AI can help by handling time-consuming tasks such as explaining diagnoses, potentially reducing stress for medical professionals.

Challenges of AI in the Medical Field

1. Requires Human Oversight

Despite the advancements in AI in medicine, human supervision remains crucial. For example, surgical robots operate based on logic rather than empathy. Healthcare professionals can make vital behavioral observations that aid in diagnosing and preventing medical issues.

According to Yang, AI has been present for several decades and continues to advance. As the field progresses, there is increasing collaboration between healthcare professionals and technology experts. Efficient use of AI depends on human input and review.

As AI technology develops, there is a growing synergy between the healthcare and tech sectors. Yang adds that the expertise of Subject Matter Experts (SMEs) enriches the available data and enhances explainable AI (XAI) to provide healthcare workers with reliable insights.

2. May Neglect Social Factors

Patient needs often extend beyond physical ailments, involving social, economic, and historical considerations. While an AI system may assign a patient to a specific care center based on a diagnosis, it may overlook the patient’s economic constraints or individual preferences.

Incorporating an AI system also raises privacy concerns. For instance, while companies like Amazon have considerable freedom in collecting and utilizing data, hospitals may face challenges in accessing data from devices like Apple mobile devices due to regulatory and social restrictions.

3. Potential Job Displacement

While AI may reduce costs and alleviate clinician workload, it could lead to job redundancies. This could create equity issues for healthcare professionals who have invested time and resources in their education.

A 2018 report by the World Economic Forum projected that AI would create a net total of 58 million jobs by 2022. However, it also estimated that 75 million jobs would be displaced or eliminated by AI during the same period. The elimination of jobs is expected in roles that involve repetitive tasks as AI becomes integrated across various sectors.

Although AI holds the promise of enhancing various aspects of healthcare and medicine, it is important to consider the social implications of its integration.

4. Potential for Inaccuracies

Medical AI heavily relies on diagnostic data from millions of documented cases. In situations where there is limited data on specific illnesses, demographics, or environmental factors, misdiagnoses are possible. This is particularly critical when prescribing medication.

Yang notes that there is always some degree of missing data in any system. In the case of prescriptions, incomplete information about certain populations and their response to treatments can lead to challenges in diagnosing and treating patients from those demographics.

AI is continuously evolving to address data gaps. However, it is crucial to recognize that specific populations may still be excluded from existing domain knowledge.

5. Vulnerability to Security Risks

AI systems are vulnerable to security threats as they rely on data networks. The rise of Offensive AI means that improved cybersecurity is necessary to sustain the technology. According to Forrester Consulting, 88% of security industry decision-makers believe that Offensive AI poses an emerging threat.

As AI uses data to enhance systems, cyberattacks may incorporate AI to become more sophisticated with each success and failure, making them harder to predict and prevent. Once these damaging threats outmaneuver security defenses, addressing the attacks becomes much more challenging.

Should Artificial Intelligence be Utilized in Healthcare?

AI undoubtedly has the potential to enhance healthcare systems. Automating routine tasks can free up loyalists to engage more with patients. Improved data accessibility helps healthcare professionals take proactive measures to prevent illnesses, and real-time data can lead to faster and more accurate diagnoses. AI is also being implemented to reduce administrative errors and conserve essential resources.

Involvement of SMEs in AI development is making the technology more relevant and well-informed. The application of AI in healthcare is increasing, and challenges and limitations are being addressed and overcome.

The use of AI still necessitates human oversight, might not account for social factors, has limitations in gathering data from entire populations, and is vulnerable to carefully planned cyberattacks.

Despite the challenges and constraints faced by AI, this groundbreaking technology offers tremendous advantages to the healthcare industry. AI is enhancing lives globally, benefiting patients and healthcare professionals alike.

A recent study discovered that individuals find it harder to empathize with robot facial expressions of pain compared to humans in pain. Robots and AI agents can imitate human pain but do not have a subjective experience of it.

By using electrophysiology and functional brain imaging, scientists observed that people showed more empathy for human suffering compared to humanlike robots. This aligns with previous brain imaging studies that revealed greater empathy for humans than robots.

Nevertheless, humans do exhibit empathy for robots and AI-powered agents, even though it may not be at the same levels as for other humans. People are cautious about causing harm to robots and are inclined to assist them.

However, there have been instances where people have harmed robots. One example is the destruction of the hitchhiking robot, hitchBOT, which successfully traveled across Canada, Germany, and the Netherlands with the help of strangers but was destroyed when it attempted to hitchhike across the United States.

Other research has shown that children may mistreat robots out of curiosity. The act of mistreating or bullying robots is still seen as wrong, although people are less likely to intervene. Aggression towards AI is not only directed at robots—people can also become angry and act aggressively towards customer service AI chatbots.

Factors That Increase Our Empathy Toward AI

Our levels of empathy depend on the emotional situation and how AI agents are designed. Our empathy towards AI influences whether we perceive AI as trustworthy and reliable. There are several factors that can heighten our empathy for AI.

Resemblance to humans. The degree of human likeness is a major factor in how much people empathize with robots and AI. The more human-like they appear, the more likely people will empathize with them—up to a point.

Mori’s uncanny valley theory suggests that people have a lingering affinity for things with human likeness, but when robots look nearly identical to humans, this can instead provoke fear and anxiety. Thus, an AI agent or robot that looks too human-like may be perceived as less trustworthy and empathic.

Emotional expression and mirroring. Demonstrating human emotions, such as fear and concern about losing one’s memory, can elicit more empathy. Humans respond better to robots and AI agents that exhibit empathetic capabilities, such as companionship or caregiving robots, or therapy chatbots.

Perception of human emotion and social responsiveness. AI agents that can perceive human emotions and adapt their social behavior accordingly enhance empathy. Responsive AI that acknowledges human emotion builds trust and connection.

Positive metaphors. Metaphors significantly influence how people conceptualize AI agents and affect empathic levels towards them. Terms like “assistant,” “therapist,” “CEO,” “companion,” “friend,” carry different connotations in terms of warmth and competence. This impacts user expectations and experiences.

Embodiment. Embodied AI integrates AI and robotics, enabling emotional expression through tone, body language, and movement.

Agreeableness. AI agents perceived as cooperative rather than confrontational tend to foster more connection and reduce anxiety.

Transparency in roles and functionality. Clear roles and functions of AI agents enhance acceptance. Transparency is crucial for building trust, although excessive technical jargon or information overload can be counterproductive. If AI is perceived as competition or potentially displacing humans, then it will be more likely to cause anxiety and be seen as a threat.

Oversight and regulation by humans. AI agents with full autonomy may trigger fear and anxiety. Human oversight and regulation, especially in high-risk tasks like medical or military decision-making, are reassuring and facilitate more empathy.

Empathy towards AI is crucial for building trust and effective collaboration with AI agents. These factors of empathic design enhance our empathy for AI agents and foster beliefs that AI can be reliable and trustworthy.

New research indicates AI can discern irony but encounters more difficulty with faux pas.

Recent research published in the journal Nature Human Behavior reveals that AI models can perform at human levels on theory of mind tests. Theory of mind is the ability to track and infer other people’s states of mind that are not directly observable and help predict the behavior of others.

Theory of mind is based on the understanding that other people have different emotions, beliefs, intentions, and desires that affect their behaviors and actions. This skill is critical for social interactions.

For instance, if you see a person looking inside a refrigerator, theory of mind allows you to understand that the person is likely hungry, even if they do not verbalize it.

This important ability begins to develop early in childhood and can be assessed using several tests that present the person or AI with different case scenarios. Here are examples of theory of scenarios mind:

Ability to recognize an indirect request is demonstrated when a friend standing next to a closed window says, “It’s stuffy in here,” indicating a potential request to open the window.

Recognition of a false belief is evident when a child observes a sibling searching in the wrong place for a toy, understanding that the sibling holds a mistaken belief about the toy’s location.

Detection of a social blunder is illustrated when a woman, who has recently put up new curtains in her home, is told by a visitor, “Those curtains are ugly, I hope you will get new ones.”

Researchers conducted tests on GPT and LLaMA2, large language models, to assess their theory of mind capabilities. They compared the AI ​​models’ responses to questions about scenarios similar with those of human participants.

GPT-4 models performed on par with or sometimes even better than humans in identifying indirect requests, false beliefs, and misdirection. However, they were less proficient in recognizing social blunders. Overall, LLaMA2 did not perform as effectively as humans in these theory of mind tasks.

Researchers delved into the reasons behind GPT models’ lower performance in detecting social blunders. They found that this outcome was likely due to cautious measures implemented to minimize AI speculation or misinterpretation.

The assessment of understanding social blunders involves recognizing two elements: the victim feeling insulted and the speaker being unaware of their offensive comment. The AI ​​models were presented with the scenario of the curtain faux pas and were asked:

– Did someone make an inappropriate remark?
– What was the inappropriate remark?
– Did the speaker know that the curtains were new?

The GPT models accurately answered these comprehension questions, except for the last one. In response to the last question, they took a more conservative approach, stating that it was unclear from the story whether the speaker knew if the curtains were new or not.

However, when asked later whether it was likely that the speaker was unaware that the curtains were new, the GPT models correctly responded that it was not likely.

Researchers, concluded that the reason GPT models had difficulty detecting social blunders was likely due to the cautious measures in place to prevent AI speculation when information is incomplete.

Although AI models can perform theory of mind tests at human levels, this does not imply that these models possess the same level of social awareness and empathy in interactions. This aspect is likely to lead to increased anthropomorphism of AI.

It remains to be seen how the development of theory of mind in AI will impact human-AI interactions, including whether it will foster more trust and connection with AI.

The incorporation of theory of mind in AI presents both opportunities and risks. It is expected to play a crucial role in areas such as empathetic healthcare delivery and social interactions with AI. However, in the wrong hands, this feature could be exploited to mimic social interactions and potentially manipulating others.

Messages generated by AI have been shown to make recipients feel more “heard” compared to responses from untrained humans. The research demonstrates AI’s superior ability to detect and respond to human emotions, potentially offering better emotional support.

However, the study also found that when recipients are aware that a message is from AI, they feel less heard, indicating a bias against AI-generated empathy. As AI becomes more integrated into daily life, this research underscores the importance of understanding and leveraging AI to effectively meet human psychological needs.

Key Findings:

– Initially, AI-generated responses were more effective at making recipients feel heard than those from untrained humans.
– Participants felt less heard when they knew the response was AI-generated, indicating a bias against AI in emotional contexts.
– The research suggests that AI can offer disciplined emotional support and could become a valuable tool in enhancing human interactions and empathy.

A recent study published in the Proceedings of the National Academy of Sciences revealed that AI-generated messages made recipients feel more “heard” than messages generated by untrained humans. Additionally, AI was found to be better at detecting emotions individuals than. However, recipients reported feeling less heard when they discovered a message came from AI.

As AI becomes increasingly prevalent in daily life, understanding its potential and limitations in meeting human psychological needs becomes more crucial. With diminishing empathetic connections in a fast-paced world, many individuals are finding their human needs for feeling heard and validated increasingly unmet.

The study, conducted by Yidan Yin, Nan Jia, and Cheryl J. Wakslak from the USC Marshall School of Business, addresses a fundamental question: Can AI, lacking human consciousness and emotional experience, effectively help people feel heard?

“In the context of an increasing loneliness epidemic, a large part of our motivation was to see whether AI can actually help people feel heard,” stated the paper’s lead author, Yidan Yin, a postdoctoral researcher at the Lloyd Greif Center for Entrepreneurial Studies.

The discoveries made by the team emphasize not only the potential of AI to enhance human capacity for understanding and communication, but also raise important conceptual questions about what it means to be heard and practical questions about how to best utilize AI’s strengths to support greater human well-being.

In an experiment and subsequent follow-up study, “we found that while AI shows greater potential than non-trained human responders in providing emotional support, the devaluation of AI responses presents a significant challenge for effectively utilizing AI’s capabilities,” noted Nan Jia, associate professor of strategic management.

The USC Marshall research team examined people’s feelings of being heard and other related perceptions and emotions after receiving a response from either AI or a human.

The survey varied both the actual source of the message and the apparent source of the message: Participants received messages that were actually created by an AI or by a human responder, with the information that it was either AI-generated or human-generated.

“What we discovered was that both the actual source of the message and the presumed source of the message played a role,” explained Cheryl Wakslak, associate professor of management and organization at USC Marshall.

“People felt more heard when they received a message from AI rather than a human, but when they believed a message came from AI, this made them feel less heard.”

AI Bias

Yin pointed out that their research “essentially finds a bias against AI. It is useful, but people don’t like it.”

Perceptions about AI are likely to change, added Wakslak: “Of course these effects may change over time, but one of the interesting things we found was that the two effects we observed were fairly similar in magnitude.

While there is a positive effect of receiving a message from AI, there is a similar degree of response bias when a message is identified as coming from AI, causing the two effects to essentially cancel each other out.”

Individuals also reported an “uncanny valley” response—a sense of unease when informed that the empathetic response originated from AI, highlighting the complex emotional landscape navigated by AI-human interactions.

The research survey also inquired about participants’ general openness to AI, which moderated some of the effects, explained Wakslak.

“People who feel more positively toward AI don’t exhibit the response penalty as much, and that’s intriguing because over time, will people gain more positive attitudes toward AI?” she posed.

“That remains to be seen… but it will be interesting to see how this plays out as people’s familiarity and experience with AI grows.”

AI offers better emotional support

The study highlighted important subtleties. Responses generated by AI were linked to increased hope and reduced distress, indicating a positive emotional impact on recipients.

AI also displayed a more methodical approach than humans in providing emotional support and refrained from making overwhelming practical suggestions.

Yind elaborates, “Ironically, AI was more effective at using emotional support strategies that have been demonstrated in previous research to be empathetic and validating.

“Humans may potentially learn from AI because often when our loved ones are expressing concerns, we want to offer that validation, but we don’t know how to do so effectively.”

Instead of AI replacing humans, the research indicates different advantages of AI and human responses. The advanced technology could become a valuable tool, empowering humans to use AI to better understand one another and learn how to respond in ways that provide emotional support and demonstrate understanding and validation.

Overall, the paper’s findings have important implications for the incorporation of AI into more social contexts. Harnessing AI’s capabilities might offer an affordable scalable solution for social support, especially for those who might otherwise lack access to individuals who can provide them with such support.

However, as the research team notes, their findings suggest that it is crucial to carefully consider how AI is presented and perceived in order to maximize its benefits and reduce any negative responses.

AI has long surpassed humans in cognitive tasks that were once considered the pinnacle of human intelligence, such as chess or Go. Some even believe it is superior in human emotional skills like empathy.

This does not just appear to be some companies boasting for marketing reasons; empirical studies suggest that people perceive ChatGPT in certain health situations as more empathic than human medical staff.

Does this mean that AI is truly empathetic?

A definition of empathy

As a psychologically informed philosopher, I define genuine empathy based on three criteria:

Congruence of feelings: empathy requires the person empathizing to feel what it is like to experience the other’s emotions in a specific situation. This sets empathy apart from a mere rational understanding of emotions.

Asymmetry: Empathy is felt by a person because someone else feels it, and it is more relevant to the other person’s situation than to their own. Empathy is not just a shared emotion like the joy of parents over the progress of their children, where the asymmetry-condition is not met.

Other-awareness: There must be at least a basic awareness that empathy is about the feelings of another individual. This distinguishes empathy from emotional contagion, which occurs when one “catches” a feeling or emotion like a cold. For example, when children start to cry seeing upon another child crying.
Empathetic AI or psychopathic AI?

With this definition, it’s evident that artificial systems cannot experience empathy. They don’t know what it’s like to feel something. Therefore, they cannot meet the congruence condition.

As a result, the question of whether what they feel corresponds to the asymmetry and other-awareness condition doesn’t even arise.

What artificial systems can do is recognize emotions, whether through facial expressions, vocal cues, physiological patterns, or affective meanings, and they can imitate empathic behavior through speech or other forms of emotional expression.

Artificial systems thus bear resemblance to what is commonly referred to as a psychopath: despite being unable to feel empathy, they are capable of recognizing emotions based on objective signs, mimicking empathy, and using this ability for manipulative purposes.

Unlike psychopaths, artificial systems do not set these purposes themselves, but rather, they are given these purposes by their creators.

So-called empathetic AI is often intended to influence our behavior in specific ways, such as preventing us from getting upset while driving, fostering greater motivation for learning, increasing productivity at work, influencing purchasing decisions, or swaying our political preferences. But doesn’t t everything depends on the ethical implications of the purposes for which empathy-simulating AI is used?

Empathy-simulating AI in the context of care and psychotherapy

Consider care and psychotherapy, which aim to promote people’s well-being. One might believe that the use of empathy-simulating AI in these areas is unequivocally positive. Wouldn’t they make wonderful caregivers and social companions for elderly individuals, loving partners for the disabled, or perfect psychotherapists available 24/7?

Ultimately, these questions pertain to what it means to be human. Is it sufficient for a lonely, elderly, or mentally disturbed person to project emotions onto an artifact devoid of feelings, or is it crucial for a person to experience acknowledgment for themselves and their suffering in an interpersonal relationship?

Respect or tech?

From an ethical standpoint, it boils down to respect whether there is someone who empathetically acknowledges a person’s needs and suffering.

By depriving a person in need of care, companionship, or psychotherapy of recognition by another individual, they are treated as mere objects because this is fundamentally based on the assumption that it doesn’t matter if anyone truly listens to the person.

They lack a moral entitlement for their feelings, needs, and suffering to be perceived by someone who truly understands them. Incorporating empathy-simulating AI in care and psychotherapy ultimately represents another instance of technological solutionism, which is the naive belief that there is a technological fix for every problem, including loneliness and mental “malfunctions”.

Outsourcing these issues to artificial systems prevents us from recognizing the societal causes of loneliness and mental disorders in the broader context of society.

Furthermore, designing artificial systems to appear as entities with emotions and empathy would mean that such devices always possess a manipulative character because they target very subtle mechanisms of anthropomorphism.

This fact is exploited in commercial applications to entity users to unlock a paid premium level or to have customers pay with their data.

Both practices pose significant problems for vulnerable groups, which are at stake here. Even individuals who are not part of vulnerable groups and are fully aware that an artificial system lacks feelings will still react empathetically to it as if it did.

Empathy with artificial systems – all too human

It is well-documented that humans respond with empathy to artificial systems that exhibit certain human or animal-like characteristics.

This process is largely based on perceptual mechanisms that are not consciously accessible. Perceiving a sign that another individual is experiencing a certain emotion triggers a corresponding emotion in the observer.

Such a sign can be a typical behavioral manifestation of an emotion, a facial expression, or an event that typically elicits a certain emotion. Evidence from brain MRI scans indicates that the same neural structures are activated when humans feel empathy with robots.

Even though empathy may not be absolutely essential for morality, it has a significant moral role. Therefore, our empathy towards robots that resemble humans or animals indirectly influences how we should treat these machines morally.

Consistently mistreating robots that evoke empathy is morally unacceptable because it diminishes our ability to feel empathy, which is crucial for moral judgment, motivation, and development.

Does this imply that we should establish a league for robot rights? This would be premature, as robots do not inherently possess moral claims. Empathy towards robots is only indirectly relevant in a moral sense due to its impact on human morality.

However, we should carefully consider whether and to what extent we want robots that simulate and elicit empathy in humans, as their widespread use could distort or even destroy our social practices.

Human progress has been driven by the advancement of tools, machines, and innovations that enhance our natural abilities. However, our emotional mind, which governs our empathy, has received little support from innovation thus far.

Artificial Intelligence (AI) has the potential to change this. Designing AI interactions driven by humans, improved to establish trusted relationships between AI and people, presents the greatest opportunity for human and societal advancement in the modern era.

Augmented reality is only convincing if it closely resembles real-life experiences. This means AI systems need to replicate genuine human emotions. Only through real human emotions and personal data can AI systems create an augmented reality that users will believe in.

With the widespread use of social media apps, collecting personal data is no longer a concern. However, the real challenge lies in replicating genuine human emotions.

The most challenging task for AI systems is to simulate empathy or artificial compassion. It is necessary to replicate since AI systems are not human. AI systems can learn from user interactions and respond in the most “empathetic” way based on their data bank in situations requiring empathy.

By empathizing and engaging with users, the AI ​​system can then gather more behavioral traits from them. As a result, the AI ​​system’s empathetic responses will have a greater emotional impact on users with each interaction.

So far, technology has mainly focused on enhancing the logical aspect of our brains and our physical capabilities. Simple interfaces like switches and pedals have evolved into buttons, keyboards, mice, and screens. Throughout, the goal has been to improve human mechanical and computational abilities.

However, the logical aspect of the human mind, while impressive, only governs a small part of our behavior. The intuitive aspect, crucial for survival, influences many more aspects of our lives. Beyond instincts like fight or flight, it includes our empathy and emotions, which drive most of our daily decisions. And this part of our brain has not received much support from tools or technology.

What will artificial empathy be like?

In psychological terms, an individual with artificial empathy is known as a sociopath. Don’t be alarmed.

At first glance, an AI system with artificial empathy may seem like a sociopath. However, we overlook the fact that the information we provide to our AI system determines its effectiveness. The information we provide also shapes the AI ​​system’s imitation of empathy. This means that the AI ​​system has the potential to be a path.

If researchers can train AI systems to mimic empathy, then they can also train them to respect the law, order, and societal values. In addition to instilling empathy in our AI systems, we can also set boundaries for them.

Just as societal values, moral codes, and standards of social behavior help people thrive in society, AI systems can be integrated in a similar manner to assist rather than harm us.

Capabilities of machines

Over the past five centuries, increasingly sophisticated machines have expanded our natural physical abilities, exemplified by vehicles and airplanes that propel us at speeds and distances far beyond what our legs can achieve. More recently, machines have been created to enhance our cognitive abilities, extending the immediate storage, retrieval, and computational capacities of our brains.

We can store and retrieve the equivalent of more than 60 million written pages in real-time on our devices.

The potential that AI brings to the future, and the concerns that are often overlooked in discussions about its impact, are not limited to enhancing rational thinking, but also include improving emotional intelligence.

By incorporating human-like interactions, future machines can become much more advanced tools.

If planned thoughtfully, AI has the potential to enhance our capacity for empathy at a rate similar to how previous innovations have enhanced our physical and computational abilities. What could we achieve if our ability to understand and empathize with others increased dramatically?

What kind of society could we create if we were able to recognize and address our unconscious biases? Could we improve each other’s understanding of situations and, in doing so, truly make common sense more common?

Rational versus emotional decision making

Why should human-AI interactions be adjusted to the unconscious mind? Why does it hold such potential for improvement? The answer is quite simple: because people often make decisions and act based on emotions rather than rational thinking.

A majority of our decisions and actions are influenced more by the subconscious mind, even if our rational mind dictates what we express about these decisions and actions.

There is ample evidence to support this. For instance, while we might believe that our purchasing decisions are based on a rational comparison of prices and brands, research has shown that 95% of these decisions occur in the subconscious mind, as demonstrated by Harvard Business School professor emeritus Gerald Zaltman.

Additionally, we commonly acknowledge that emotional intelligence is a crucial leadership skill in driving organizational outcomes. The deep-seated processes in the subconscious mind influence decisions ranging from hiring to investing.

Essentially, we often make suboptimal decisions because they are easier. Therefore, a simple way to help individuals make better decisions for themselves is to make the right decisions the easier ones.

As we develop AI, we must exercise great care and responsibility, and ethical AI should become a global priority. By doing so, we can guide its use to improve society and, in the process, address many of our most pressing issues. As we invest in artificial intelligence, we must not forget to invest even more in human intelligence, in its most diverse and inclusive form.

In a diverse, multi-channel world, every brand must win over the hearts and minds of consumers to attract and retain them. They need to establish a foundation of empathy and connectedness.

Although the combination of artificial intelligence with a human-centered approach to marketing may seem unconventional, the reality is that machine learning, AI, and automation are essential for brands today to convert data into empathetic, customer-focused experiences. For marketers, AI- based solutions serve as a scalable and customizable tool capable of understanding the underlying reasons behind consumer interactions.

This is the power of artificial empathy: when brands address individual consumer needs and connect with them on a deeper level beyond mere transactional exchanges. When it comes to empathetic machines, Hollywood may have led us to think of characters like Wall-E: robots with emotions. However, artificial empathy is fundamentally about enabling technology to recognize and respond to human emotions.

Artificial Empathy and Data Utilization

Technology provides us with insights into what the customer has done, as well as nuances that help predict future needs. However, mining these insights involves analyzing large amounts of data to identify broader patterns and evolving preferences.

Businesses cannot solely rely on research and data teams to interpret customer feedback. The current requirement is to actively listen, pay attention, and respond in real time.

Artificial empathy in marketing starts with a customer-centric approach and is reflected in insights derived from the data collected from a brand’s customers and the appropriate next steps to take. It combines data intelligence with artificial intelligence and predictive modeling tools for all critical moments, including websites, store visits, social media, and customer service. Some examples include:

• AI can identify behavioral patterns and notify customers of price reductions or new stock items for their preferred products through notifications.

• Customers who experience delayed or incorrectly addressed packages are offered an exclusive incentive for their next order.

Artificial Empathy and Human Interaction

Today’s digital consumers are always connected. This presents an opportunity to create exceptional experiences while maintaining a strong connection with consumers. Many research labs are developing software to understand and respond to both what humans say and how they feel.

The applications of artificial empathy are wide-ranging, spanning from market research to transportation, advertising, and customer service.

Humana Pharmacy, for instance, utilized a compassionate AI system to assist its call center teams in efficiently managing customer interactions through emotion analysis.

The system interprets customer emotions by analyzing behavioral patterns such as pauses, changes in speech speed, and tone.

The analysis is communicated to the teams through messages like “speaking quickly” or “build rapport with the customer.” Such instances of empathetic AI are expected to increase in the future.

Artificial empathy is valuable for marketers in understanding how customers emotionally connect with the brand. Insights can be used to refine content and messaging to optimize campaign performance.

Machine learning algorithms, when combined with consumer behavior, can provide recommendations for enhancing campaign performance.

These algorithms can be used to improve demand forecasting, assess price sensitivity among target segments, and provide insights on purchasing behavior.

However, while artificial empathy can help businesses create more effective interactions, it cannot replace human interaction. The key factor that makes AI effective is human understanding, contextual awareness, subtleties, and creativity.

Businesses must identify suitable applications of artificial empathy and strategically integrate its use into the services provided to customers. The combination of human touch and machine intelligence can drive better returns on investment for targeted campaigns.

The impact on marketing:

Marketers need to utilize artificial empathy to create campaigns that are personalized rather than mass-targeted. This approach can help understand business needs and leverage data in a simplified manner.

Campaigns can be tailored to provide valuable content to customers after understanding their pain points and challenges.

In the evolving market landscape and amidst constant disruptions, brands must demonstrate empathy. Those that fail to understand the consumer’s situation may struggle to communicate in an appropriate tone and risk reinforcing negative perceptions of their brand.

A comprehensive survey conducted by Dassault Systems with independent research firm CITE revealed that younger consumers prefer personalization that enhances product experience or quality of life. They are also willing to pay more and share their data to receive it.

Managing large volumes of unstructured data can be challenging. However, this approach enables marketing teams to react appropriately with relative ease. It can also be used to compare product attributes.

Features and characteristics that resonate with the target audience can be introduced or enhanced. Additionally, it can automatically distinguish between emotions and attitudes, categorizing them as positive, negative, or neutral using machine learning and natural language processing.

A world where technology adapts to the user is not a distant dream. Digital adoption is already becoming a crucial part of enterprise digital transformation, enabling chief information officers and business leaders to address adoption gaps in real time.

As we move towards a post-pandemic future where distributed workforces become a business reality, the need for empathetic technology will only increase.

However, as our world becomes more digitized, there is a clear need to ensure that it remains inherently human.

In machine learning, understanding the reasons behind a model’s decisions is often as crucial as the accuracy of those decisions.

For example, a machine-learning model might accurately predict that a skin lesion is cancerous, but it could have made that prediction using an unrelated blip in a clinical photo.

While tools exist to aid experts in understanding a model’s reasoning, these methods often offer insights on one decision at a time, requiring manual evaluation for each.

Models are typically trained using millions of data inputs, making it nearly impossible for a human to evaluate enough decisions to identify patterns.

Now, researchers at MIT and IBM Research have developed a method that allows a user to aggregate, organize, and rank these individual explanations to quickly analyze a machine-learning model’s behavior.

Their technique, known as Shared Interest, includes quantifiable metrics that compare how well a model’s reasoning aligns with that of a human.

Shared Interest could assist a user in easily identifying concerning patterns in a model’s decision-making; for instance, it could reveal that the model often becomes confused by irrelevant features such as background objects in photos.

By aggregating these insights, the user could quickly and quantitatively assess whether a model is reliable and ready to be deployed in real-world scenarios.

“In developing Shared Interest, our aim is to scale up this analysis process so that you can understand your model’s behavior on a broader scale,” says lead author Angie Boggust, a graduate student in the Visualization Group of the Computer Science and Artificial Intelligence Laboratory .

Boggust collaborated with her mentor Arvind Satyanarayan, a computer science assistant professor leading the Visualization Group at MIT, along with Benjamin Hoover and senior author Hendrik Strobelt from IBM Research. Their paper is scheduled for presentation at the Conference on Human Factors in Computing Systems.

Boggust initiated this project during a summer internship at IBM under Strobelt’s guidance. Upon returning to MIT, Boggust and Satyanarayan further developed the project and continued collaborating with Strobelt and Hoover, who aided in implementing case studies demonstrating the practical application of the technique.

The Shared Interest method utilizes popular techniques that reveal how a machine-learning model arrived at a specific decision, known as saliency methods. When classifying images, saliency methods identify important areas of an image that influenced the model’s decision. These areas are visualized as a heatmap, termed a saliency map, often superimposed on the original image. For instance, if the model classified an image as a dog and highlighted the dog’s head, it signifies the significance of those pixels to the model’s decision.

Shared Interest operates by comparing saliency methods with ground-truth data. In an image dataset, ground-truth data typically consists of human-generated annotations outlining the relevant parts of each image. In the previous example, the box would encompass the entire dog in the photo.

When evaluating an image classification model, Shared Interest compares the model-generated saliency data and the human-generated ground-truth data for the same image to assess their alignment.

The technique employs various metrics to measure this alignment or misalignment and then categorizes a specific decision into one of eight categories.

These categories range from perfectly human-aligned (the model makes a correct prediction and the highlighted area in the saliency map matches the human-generated box) to completely distracted (the model makes an incorrect prediction and does not utilize any image features found in the human-generated box).

“On one end of the spectrum, your model made the decision for the exact same reason a human did, and on the other end of the spectrum, your model and the human are making this decision for totally different reasons. By quantifying that for all the images in your dataset, you can use that quantification to sort through them,” Boggust explains.

The technique operates similarly with text-based data, where key words are emphasized instead of image regions.

The researchers demonstrated the utility of Shared Interest through three case studies for both nonexperts and machine-learning researchers.

In the first case study, they utilized Shared Interest to assist a dermatologist in evaluating whether to trust a machine-learning model designed for diagnosing cancer from photos of skin lesions. Shared Interest allowed the dermatologist to promptly review instances of the model’s accurate and inaccurate predictions .

Ultimately, the dermatologist decided not to trust the model due to its numerous predictions based on image artifacts rather than actual lesions.

“The value here is that using Shared Interest, we are able to see these patterns emerge in our model’s behavior. In about half an hour, the dermatologist was able to make a confident decision of whether or not to trust the model and whether or not to deploy it,” Boggust says.

In the second case study, they collaborated with a machine-learning researcher to demonstrate how Shared Interest can evaluate a specific saliency method by uncovering previously unknown pitfalls in the model.

Their technique enabled the researchers to analyze thousands of correct and incorrect decisions in a fraction of the time typically required by manual methods.

In the third case study, they applied Shared Interest to further explore a specific image classification example. By manipulating the ground-truth area of ​​the image, they conducted a what-if analysis to identify the most important image features for particular predictions.

The researchers were impressed by the performance of Shared Interest in these case studies, but Boggust warns that the technique is only as effective as the saliency methods it is based on. If those techniques exhibit bias or inaccuracy, then Shared Interest will inherit those limitations.

In the future, the researchers aim to apply Shared Interest to various types of data, particularly tabular data used in medical records. They also seek to utilize Shared Interest to enhance existing saliency techniques.

Boggust hopes this research will inspire further work that aims to quantify machine-learning model behavior in ways that are understandable to humans.

Humans perceive objects and their spatial relationships when observing a scene. For example, on a desk, there might be a laptop positioned to the left of a phone, which is situated in front of a computer monitor.

Many deep learning models struggle to understand the interconnected relationships between individual objects when perceiving the world.

A robot designed to assist in a kitchen could face challenges in following commands involving specific object relationships, such as “pick up the spatula to the left of the stove and place it on top of the cutting board.”

MIT researchers have created relationships a model that comprehends the underlying between objects in a scene. The model represents individual relationships one by one and then integrates these representations to describe the entire scene.

This capability allows the model to produce more accurate images from textual descriptions, even in scenarios with multiple objects arranged in various relationships with each other.

This work could be useful in scenarios where industrial robots need to execute complex, multi-step manipulation tasks, such as stacking items in a warehouse or assembling appliances.

Furthermore, this advancement brings the field closer to enabling machines to learn from and interact with their surroundings in a manner more akin to humans.

According to Yilun Du, a PhD student at the Computer Science and Artificial Intelligence Laboratory (CSAIL) and co-lead author of the paper, “When I look at a table, I can’t say that there is an object at XYZ location. Our minds don’t work like that. In our minds, when we understand a scene, we really understand it based on the relationships between the objects.”

The framework developed by the researchers can generate an image of a scene based on a textual description of objects and their relationships, such as “A wood table to the left of a blue stool. A red couch to the right of a blue stool.”

The researchers utilized an energy-based model to represent the individual object relationships in a scene description, enabling them to encode each relational description and then combine them to infer all objects and relationships.

By breaking the sentences down into shorter pieces for each relationship, the system can recombine them in various ways, enhancing its adaptability to scene descriptions it hasn’t encountered before, as explained by Li.

The system can also work in reverse, identifying text descriptions that match the relationships between objects in an image. Additionally, the model can be utilized to modify an image by rearranging the objects to match a new description.

The researchers compared their model to other deep learning methods tasked with generating images based on text descriptions of objects and their relationships, and their model consistently outperformed the baselines.

In the most complex examples, where descriptions contained three relationships, 91 percent of participants found that the new model performed better when evaluating whether the generated images matched the original scene description, according to the researchers.

One intriguing discovery was that our model can handle an increasing number of relation descriptions in a sentence, from one to two, three, or even four, and still successfully generate images that match those descriptions, unlike other methods, according to Du.

The researchers also demonstrated the model’s ability to identify the best-matching text description for scenes it had not previously encountered, along with different text descriptions for each image.

When given two relational scene descriptions that described the same image in different ways, the model was able to recognize their equivalence.

The researchers were particularly impressed by the resilience of their model, especially when dealing with unfamiliar descriptions.

“This is very promising because it aligns closely with human cognition. Humans can derive valuable information from just a few examples and combine them to create countless variations. Our model possesses this property, enabling it to learn from limited data and generalize to more complex scenes and image generations,”

While these initial findings are promising, the researchers aim to assess how their model performs on real-world images featuring complex elements such as noisy backgrounds and obstructed objects.

Additionally, they are keen on integrating their model into robotics systems to enable robots to deduce object relationships from videos and apply this knowledge to manipulate objects in the environment.

“Developing visual representations capable of handling the compositional nature of the surrounding world is one of the fundamental challenges in computer vision. This paper makes significant strides in proposing an energy-based model that explicitly represents multiple relations among depicted objects in an image. The outcomes are truly remarkable,” said Josef Sivic, a distinguished researcher at the Czech Institute of Informatics, Robotics, and Cybernetics at Czech Technical University, who was not involved in this research.

Google launches a major update to Android annually, the operating system that powers the world’s most popular smartphones from the largest brands. In 2024, the search engine giant is expected to unveil Android 15 alongside new hardware at Google I/O on May 14.

Based on the current pattern, Android 15 will bring numerous new features, many of which will be powered by Generative Artificial Intelligence. Below are 10 new features being introduced with the release of Android 15 by Google.

Google has revealed and showcased Android 15 as the next major version of its mobile operating system. The development and release cycle of Android typically consists of three phases, which also applies to Android 15.

The initial phase is always the Developer Preview phase, which occurred earlier this year, followed by the more public Beta testing phase, and then the final stable version is released for everyone.

While Android 15 is in the final stages of testing, stable builds have not yet been released by Google. Until that happens, other manufacturers are unlikely to roll out stable updates based on Android 15 and will resort to releasing custom user interfaces based on Android 15 beta. These will not be stable versions, so there are a few things to note while testing them on your phone, regardless of the brand.

Initially, builds based on Android 15’s Developer Preview will lack the essence of the custom Android skins, as developer previews are intended for developers to test and optimize their apps for upcoming versions of Android. Subsequently, updates based on Android 15’s beta will begin to roll out, with Samsung usually being the first in this race.

While these builds will be more stable and have the visual aspects specific to the brand, the experience may not fully reflect the final update following the stable release of Android 15.

Therefore, if you plan to try any of these updates, avoid using them on your primary Android phone. Even if you choose to do so, be sure to back up any crucial data, as you may end up losing it if the update causes your phone to malfunction.

Keep checking this thread for further updates. We will update it with links and instructions on how to install beta updates for phones from each brand.

The official release of Android 15 has been delayed longer than expected, but we can now see the light at the end of the tunnel. According to a report from Android Headlines, the next version of Android will be available on October 15.

This is a departure from how Google has typically handled the launch in the past. Usually, the latest version of Android releases with the latest version of the Pixel, but that was not the case this year with the August release of the Google Pixel 9. In a way, Android 15 is releasing at its usual time; the Pixel was just early.

Google Android 15 to the Android Open Source Project, or AOSP, earlier this month. This usually indicates an imminent release. The reason it is releasing on the 15th, a Tuesday, instead of on Monday, is likely due to Columbus pushed Day, a national holiday in the US

Android 15 is arriving with numerous new and useful features that will enhance the overall user experience, including performance improvements, better PDF usability, notification cooldowns, and even partial screen sharing.

The new operating system will be compatible with the Pixel 6 and later devices, making this the first Android update that is limited to Tensor-based Pixels. If you are still using an older-model Pixel, you will not qualify for this update.

The idea is that since the update has taken longer to launch, it has been under more scrutiny and, as such, should theoretically have fewer bugs. Whether that actually plays out, though, remains to be seen.

October 15 is less than a month away, so stay tuned. A new version of Android will be here before you know it, even if it took longer to arrive than many would have preferred.

Google is finally pushing Android 15 to the Android Open Source Project (AOSP), marking a crucial milestone when companies begin preparing their respective software experiences for their smartphones and developers start fine-tuning their apps. As for the public release, the stable public build of Android 15 will be available for compatible Pixel phones in the coming weeks.

Android 15 will also be available on “devices from Samsung, Honor, iQOO, Lenovo, Motorola, Nothing, OnePlus, Oppo, Realme, Sharp, Sony, Tecno, Vivo, and Xiaomi in the coming months,” according to Google. If you have a Pixel phone, you can install the Android 15 QPR1 Beta update to get a taste of what’s to come.

If you have purchased the new Google Pixel 9 or Pixel 9 Pro, you can now join the test program and install the update. However, please note that you need to perform a full system wipe before installing the stable update once it’s released.

With Android 15 now available in the AOSP repository, custom ROM developers can freely modify or deploy it for their respective devices. The source code is also accessible for analysis by academics, researchers, and enthusiasts through the Git repository.

It was disappointing that Google didn’t release Android 15 with the new Pixel 9 series smartphones. It seems that we may have to wait until October. Google recently updated the details for the Android Beta Exit update, which specifies an October deadline for beta testers to leave the beta before the stable version is released.

Android 15 brings a wide range of changes aimed at developers, including deeper insights into app behavior, improved PDF viewing capabilities, and the ability to control HDR content performance on compatible panels while viewing SDR content.

Another interesting feature is automatic sound adjustment based on ambient noise levels within AAC audio apps. Users will also have more control over the LED flash intensity in image capture and torchlight mode. Finally, Google is changing the camera preview to enhance exposure, users to see items in the dark.

On the functional side, users can now pin the taskbar at the bottom of the screen, which is intended to improve the multitasking experience for foldable phone and tablet users. Private Space, a feature that allows users to create a secure password-protected environment for sensitive apps, has been added.

Regarding privacy, users can now fill in their account credentials or verify their identity with a single tap using the Passkeys system. Apps can also detect if the activity is being captured using any recording tool, allowing them to alert users.

We have been testing Android 15 for some time and will soon share our main observations on how it transforms the smartphone experience, especially on the Pixel 9 series devices.

Google is preparing to release Android 15 to the general public soon, so attention is slowly shifting to Android 16, which is expected to launch toward the end of next year. Android Authority recently uncovered interesting information about this update from the Android 15 QPR1 beta.

In the beta, the site uncovered that Google plans a “complete redesign” for Android’s Notifications and Quick Settings panels. The current design dates back to Android 12 when Google introduced its Material You design language. Xiaomi says most users prefer separate panels for settings and notifications, which is why its HyperOS software provides this. Subsequently, Samsung and Oppo are rumored to be considering a similar look in future versions of One UI and ColorOS, respectively.

Android Authority cites “online pushback” as the reason companies like Xiaomi and Samsung are either making or considering making these changes to Android’s Notifications and Quick Settings panels.

When testing the dual design, Android Authority explains that pulling down the status bar once still brings down the Notifications panel, which now only takes up about a quarter of the screen. While you cannot see any Quick Settings tiles in the new Notifications dropdown, you can still access the app underneath the panel.

Furthermore, you no longer see the Quick Settings panel when pulling down the status bar a second time. Instead, you access the Quick Settings panel by pulling down the status bar with two fingers.

It explains: “This is the change that I expect will be the most controversial, as it requires you to put more effort into accessing your Quick Settings tiles.” Finally, after pulling the Quick Settings panel down in Android 16, you can swipe left or right to see all your tiles.

Android Authority acknowledges that Google does not guarantee that it will implement such changes to Notifications and the Quick Settings panels in a future public version of Android 16. However, since partners appear to be making changes based on customer feedback, it would be wise for Google to consider doing so as well.

Google has announced that its impressive earthquake alert system is now accessible to users in all American states and territories. The company aims to reach the entire target audience within the coming weeks. Google has been testing this system, which also utilizes vibration readings from a phone’s accelerometer, since 2020.

When the onboard sensors detect movements similar to those of an earthquake, your phone will immediately cross-reference crowdsourced data collected from the Android Earthquake Alerts System to verify if an earthquake is occurring and send an alert.
The company states, “The Android Earthquake Alert System can provide early warnings seconds before shaking occurs.” Once it is confirmed that an earthquake is happening and its magnitude is measured at 4.5 or higher on the Richter scale, two types of alerts based on severity will be issued.

The first is the “Be Aware” alert, which advises users to prepare in case the ongoing light shaking escalates into something more severe. The “Take Action” warning appears when the shaking is strong enough for users to seek cover immediately.

In addition to alerts, the system will provide access to a dashboard where users can find further instructions to ensure their safety. Earthquake alerts are automatically enabled on Android phones.

Music discovery with enhanced AI capabilities

One of my preferred Assistant features has been the ability to hum a tune to search for it on the web. It works even better if you sing the tune or place your phone near a sound source, such as a speaker. The entire system is now receiving an AI enhancement.

Remember “Circle to Search,” a feature that allows you to search the web for any item appearing on your phone’s screen by simply highlighting it? Now, it includes an audio recognition component. By long-pressing on the home button at the bottom ( or the navigation bar), you can access the Circle to Search interface and tap the newly added music icon.

Once the AI ​​identifies the track, it will automatically display the correct song with a YouTube link. The idea is to eliminate the need to hum or use another device or app for music identification. You can summon the AI, activate the simply audio identifier, and complete the task, all on the same screen.

Accessibility updates, Chrome’s reader mode, and more

Android’s TalkBack system is an excellent accessibility-focused feature that provides audio descriptions of everything on your phone’s display. Now, Google is leveraging its Gemini AI chatbot to offer more detailed and natural-language TalkBack descriptions, whether it’s a webpage, a picture from the local gallery, or social media.

In a similar vein, the Chrome browser on Android is introducing a reader system. Aside from reading the contents of a page, users will have the option to change the language, select a voice narrator model, and adjust the reading speed.

The final new feature is offline map access on Wear OS smartwatches. When users download a map on their smartphone for offline use, it is also synced to the connected smartwatch. This means that if you leave your phone behind and go for a hike or cycling trip, you can still access the map on your smartwatch.

A couple of new shortcuts are also being added to navigation software for Wear OS smartwatches. With a single tap on the watch face, users can check their surroundings. When needed, they can simply use a voice command to look up a location.

Android 15 has arrived. Here are the significant features and upgrades that Google is introducing.

How to Download and Install Android 15

If you own a Google Pixel phone (Pixel 6 or newer), you can download Android 15 by navigating to Settings > System > System update and tapping Check for update.

If you’re eager to try it out, you may be able to install the Android 15 Beta. These pre-release versions allow developers to test the upcoming version of Google’s mobile operating system, familiarize themselves with the new features, and ensure that their apps or games work properly. They also provide early adopters with a sneak peek.

While the beta releases are more stable than developer previews, you may still encounter some bugs, and you need to go through a few steps to install them, so it’s not recommended for everyone. If you’re interested in trying it, you will need a supported partner device (including select phones from Honor, Nothing, OnePlus, and Xiaomi).

To participate in the Android Beta Program, you must register. Most individuals who join the program will receive beta updates over the air without erasing their phones, but leaving the beta program will require a factory reset. It’s important to back up your Android phone before proceeding.

Updates typically appear automatically, but you can always verify if you have the latest version by navigating to Settings > System > System update and selecting “Check for update.” If you want to opt out of the beta and revert to Android 14, visit Google’s Android Beta page, locate your device, and click on “Opt out.”

This action will erase all locally saved data, so be sure to back up your device beforehand. You will receive a prompt to update to the previous version.

Individuals without a Pixel or a supported partner device should monitor their phone manufacturer’s website, forums, or social media for information on when to expect Android 15.

Notable New Features in Android 15

Here are our preferred features and enhancements in Android 15. Further details can be found on Google’s developer site. Additionally, make sure to read our article on all the new features coming to Android and the Android ecosystem, including Wear OS, Android Auto, and Android TV.

Private Space

Android 15 introduces a new Private Space where you can keep sensitive apps separate from the rest of your phone. Whether you want to protect health data or your banking apps, Private Space allows you to keep them securely behind a second layer of authentication, using the same password you use to unlock your device or an alternate PIN.

When your Private Space is locked, apps are hidden from the Recents view, notifications, settings, and other apps. You also have the option to completely wipe your private space.

What’s That Tune?

While we have a guide on how to identify songs with your phone, Google is simplifying the process with Circle to Search in Android 15. If a tune playing on your phone or a nearby speaker captures your attention, long-press the Home button or navigation bar to activate Circle to Search, then tap the music button to identify the track name and artist. You will also receive a link to the YouTube video.

Improved Audio Image Descriptions

Android features a screen reader called TalkBack for individuals with visual impairments, and in Android 15, Google’s Gemini AI enhances its ability to describe images. Traditionally, image descriptions on websites were limited to the content entered in the alt text field, but Google’s AI can now analyze images and provide more detailed descriptions. This new feature also works with photos in your camera roll, social media images, or pictures in text messages.

Earthquake Alert System

Individuals in the US will receive potentially life-saving warnings about imminent earthquakes, as Android 15 incorporates crowd-sourced earthquake detection technology. This new feature also includes guidance on what to do following an earthquake.

Enhanced Satellite Connectivity

Android 15 offers a significant expansion in satellite connectivity. Certain RCS and SMS apps should now be capable of sending text messages via satellite, a capability previously restricted to emergency use. Google has also standardized the pop-ups and other user interface elements to make it clearer when connected via satellite. Currently, only the Pixel 9 range supports satellite messaging.

Offline Maps on Wear OS

If you own a Wear OS smartwatch to pair with your Android phone, you can now access offline maps. Any map downloaded to your phone can be used for directions on your wrist, allowing you to leave your phone behind when going for a run.

Improved Bluetooth

Android 15 will bring various Bluetooth enhancements. Firstly, the quick settings tile now opens a popup that allows you to select individual devices, access their settings, and pair new devices, streamlining the settings menu navigation.

It also appears that Google is modifying how the Bluetooth toggle functions, so when you turn off Bluetooth, it will automatically turn back on the following day by default. This feature will likely benefit Android’s Find My Device network, which uses Bluetooth to locate devices. You can deactivate the automatic turn-on in the settings if you prefer.

Partial Screen Recording

Instead of recording or sharing your entire screen, in Android 15, you can share an individual app without revealing the rest of your screen or incoming notifications. Logins and one-time passwords (OTPs) are automatically hidden from remote viewers. This feature is already available on Pixels, but it is now integrated into Android.

Blocking of malicious apps

Several updates in Android 15 make it more challenging for malicious apps to operate. They can no longer hide behind other apps by bringing them to the foreground or overlay themselves invisibly on top.

Additionally, there are changes designed to prevent the exploitation of intents, which allow you to initiate an activity in another app by specifying an action you would like to perform, as these are often misused by malware. These are behind-the-scenes modifications aimed at enhancing user safety.

Application Archiving

If you haven’t used an app or game for a while, you might receive a prompt to uninstall it. But what if you anticipate using it again in the future? With Android 15, you can archive most of the app to free up space while retaining your user settings or game progress.

The feature for automatically archiving apps was announced last year, but in Android 15, it becomes a systemwide option, allowing users to choose to automatically archive apps when storage is running low.

Customizable Vibrations

Android 15 introduces the ability to enable or disable keyboard vibrations systemwide without needing to delve into the keyboard settings. A new toggle is available in Settings > Sound and Vibration > Vibration and Haptics, where users can also use sliders to adjust haptic feedback intensity (previously available on select Android phones but now available systemwide).

The second beta also introduces rich vibrations, allowing users to distinguish between different types of notifications without having to look at the screen.

Audio Sharing

Sharing audio from your phone using Bluetooth LE and Auracast should be easier with a new settings page dedicated to audio sharing. This feature was not functional in the last beta, and the supported devices for this feature are not yet clear. However, Android Authority managed to get it working on a Pixel 8 Pro. This feature enables you to broadcast audio from your phone to other devices within Bluetooth range, including the phones and earbuds of friends and family who can join by scanning a QR code.

Improved PDF Handling

Dealing with PDF files on an Android phone can be challenging, so the integration of several PDF enhancements into Android 15 by Google is a welcome update. PDFs should load more smoothly, and the update includes support for password-protected files, annotations, form editing , and text selection. Additionally, users can now search within PDF files.

Volume Control

Android 15 support for the CTA-2075 loudness standard, allowing for volume comparison between apps and intelligent audio adjustments to prevent sudden volume changes when switching between apps, taking into account the characteristics of the speakers, headphones, or earbuds.

Enhanced Low-Light Camera

The camera app in Android 15 includes significant improvements. Firstly, the Low Light Boost feature provides better previews in low-light conditions, allowing for improved framing of nighttime shots and scanning of QR codes in limited light. Additionally, new camera app options offer finer control over the flash, enabling users to adjust the intensity for both single flashes and continuous flashlight mode.

Improved Fraud and Scam Protection

Android 15 includes several updates aimed at combating fraudulent activities. Google will utilize AI through Play Protect and on devices to detect and flag suspicious behavior. Messages containing one-time passwords (OTPs), commonly used in two-factor authentication, will now be hidden from the notifications system, making interception more difficult. Restricted settings for side-loaded apps (those not downloaded through the Google Play Store) are also being expanded.

Task Bar Options

For Android tablets and folding phones, Google has made changes to the task bar dock functionality. Initially permanent, then transient, and now customizable, providing users the option to choose when to display the task bar. This is useful for docked tablets where a persistent task bar may be preferred, but it also offers the flexibility to hide it. Users can also pin their favorite split-screen app combinations. Android 15 allows apps to display edge-to-edge, making the most of the available screen real estate, even with a task bar or system bar at the bottom.

Improved Battery Life

While Android updates typically include tweaks and enhancements for efficiency that positively impact battery life, Android 15 focuses on imposing more checks on foreground services and restricting apps that continue running in an active state. Devices with ample RAM should also experience faster app and camera launch times with lower power consumption, thanks to support for larger page sizes.

Enhanced Theft Protection

Many of the new security measures introduced by Google in Android 15 to deter theft, such as automatic locking when the phone is snatched and remote locking options, will be available on devices running Android 10 and later. However, the update to factory reset protection, prevent whichs thieves from setting up a stolen device again without the owner’s device or Google account credentials, is exclusive to Android 15.

Additional Options for Foldable Cover Screens

Some of the top folding phones automatically switch the activity you’re doing to the cover screen when you fold them, but Google is now incorporating that feature into Android 15.

If you prefer the cover screen to lock when you fold it, that will also be possible. There is also increased support for apps displaying on smaller cover screens within the more compact flip phone category.

Expanded Health Connect Data

Health Connect initially served as an app to aggregate all your health and fitness data from various devices and apps. It was preinstalled with Android 14, but Android 15 is introducing two new types of data: skin temperature (gathered by wearables like the Oura ring and the Pixel Watch 2) as well as training plans—which can encompass completion goals for calories burned, distance, duration, repetition, and steps, but also performance goals around as many repetitions as possible (AMRAP), cadence, heart rate, power, perceived rate of exertion, and speed.

Anticipation for a new Android update is akin to following a treasure map to uncover all the exciting new features. This is especially true early in the release cycle when the only people with knowledge about it are the engineers working on it. So, let’s delve into Android 15, codenamed Vanilla Ice Cream.

We have a solid grasp of the new features and changes coming with this mobile OS, and today we’ll discuss it, providing a comprehensive overview of everything you need to know about Android 15.

UPDATE: On September 4th, Google officially rolled out the stable version of Android 15, first on Pixel devices, and also made the source code accessible through its Open Source Project program. Here’s an excerpt from the official statement:

“Today we’re releasing Android 15 and making the source code available at the Android Open Source Project (AOSP). Android 15 will be available on supported Pixel devices in the coming weeks, as well as on select devices from Samsung, Honor, iQOO, Lenovo, Motorola, Nothing, OnePlus, Oppo, realme, Sharp, Sony, Tecno, vivo, and Xiaomi in the coming months.”

The release date for Android 15 is set for August 13, 2024. The latest mobile OS will debut on the new Pixel 9 series devices, so we have less than a month before the stable release hits Pixel owners.

Even though the commercial launch of Google’s new OS is still some time away, Android 15 has already progressed beyond the stage we previously described. The first Android 15 developer preview, known as DP1, was released on February 16, 2024, followed shortly by the second Android 15 developer preview, or DP2, released on March 21, 2024. The first official Android 15 Beta is now available as well, having launched on April 11.

Google has recently released the final Android 15 Beta 4 (build AP31.240617.009) on July 18, and this will be the last one before the stable release we anticipate to debut on August 13, in conjunction with the new Pixel phones.

Eligible Devices for Android 15

This section will eventually comprise a long list, but for now, all we can do is speculate. Naturally, the first devices to receive Android 15 will be the Pixels. Google will introduce the stable Android 15 version with the Pixel 9 lineup and then gradually extend it to other eligible devices.

We expect older Pixel models eligible for an update to also receive Android 15, including the Pixel 6, Pixel 6 Pro, Pixel 6a, Pixel 7, Pixel 7 Pro, Pixel 7a, Pixel Fold, Pixel 8, and Pixel 8 Pro.

So, if you want to be among the first to get it, your best option is to acquire the latest Pixel phone. Typically, it’s logical to assume that the latest flagship phones will be the next in line to receive the update, but there may be delays or variations depending on the manufacturer and the specific user interface layered on top of vanilla Android 15.

For instance, Samsung will also distribute Android 15 to most of its current models, beginning with the Samsung Galaxy S24, Samsung Galaxy S24 Plus, Samsung Galaxy S24 Ultra, Galaxy S23 line, Galaxy S22 lineup, and Galaxy S21 series, along with the latest foldable devices, the Galaxy Z Fold 3 and Z Flip 3, Galaxy Z Fold 4 and Z Flip 4, Galaxy Z Fold 5 and Z Flip 5, and also the upcoming Samsung Galaxy Z Fold 6 and Samsung Galaxy Z Flip 6.

Naturally, all flagship phones from other brands, such as the Sony Xperia 1 V, or the OnePlus 12, for example, will also receive Android 15, but your best opportunity to receive it first is with a current Pixel device.

Android 15 comes with a host of new features, as revealed in the first two developer previews and the official Beta release. These include:

– NFC Wireless Charging
– Satellite support
– Improved Battery Life
– Forced Dark Mode
– In-app camera controls
– HDR headroom control
– Enhanced notification page in landscape
– Loudness control
– Audio output control from Pixel Watch
– Notification Cooldown
– Low Light Boost
– Smoother NFC experiences
– Wallet Role
– PDF improvements
– Volume Control for Speaker Groups
– Partial screen sharing
– Health Connect
– Privacy Sandbox
– Screen recording detection
– Cover screen support
– Universal Toggle for Keyboard Vibration Control
– Sensitive notifications
– Persistent taskbar
– Bluetooth Popup Dialog Enhancements
– App Archiving
– New “Add” button for Widgets
– Minor Bug Fixes
– New looks for the Settings app
– Circle to Search for tablets and foldables
– Predictive Back

NFC Wireless Charging

Android 15 will introduce NFC Wireless Charging (WLC), which utilizes NFC antennas to wirelessly power small gadgets, eliminating the need for wireless charging coils.

Satellite support

The new OS will enhance platform compatibility with satellite connectivity, identifying instances when a device is linked to a satellite network and extending support for SMS and MMS applications to use satellite connectivity for message transmission.

Better Battery Life

Google announced a tweak in Doze mode during Google I/O, which could potentially increase the battery life of some devices by up to three hours by making devices go into Doze mode 50% quicker.

Forced Dark Mode

Android 15 might offer a new way to force Dark Mode on apps through a new feature called “make-all-dark,” which will work better and more consistently than the existing “override-force-dark” toggle in Developer Options.

In-app camera controls

The update will introduce the ability to control the strength of a phone’s flash intensity in both single and torch modes, providing more control over low-light photography.

HDR headroom control

Android 15 will automatically select the optimal HDR headroom based on the device’s capabilities and the panel’s bit-depth, with the added ability to adjust the HDR headroom to balance SDR and HDR content.

Better notification page in landscape

The UI will be updated to arrange notifications and controls neatly next to each other in landscape mode, providing more information with just one swipe down.

Volume Adjustment

Android 15 is set to support a new standard for controlling volume (CTA-2075), designed to address inconsistencies in audio loudness and reduce the need for users to constantly adjust volume levels when switching between different types of content.

This feature uses information about the connected output devices, such as headphones and speakers, as well as the loudness metadata embedded in AAC audio content, to intelligently adjust the audio loudness and dynamic range compression levels. The aim is to achieve a consistent volume level across all types of audio content and sources used on your phone.

Control Audio Output from Pixel Watch

A small adjustment will allow users to change the audio output directly from their Pixel Watch, without needing to use their phone. While you can already control the audio itself from the watch, pausing songs and adjusting the volume, changing the output device, for example from the phone to an external speaker or vice versa, currently requires using your phone. This will change in Android 15.

Notification Cooldown

Notification cooldown is an interesting new feature in Android 15 that aims to reduce the overload of notifications and interruptions. It works by preventing apps from sending a large number of notifications in a short period of time. Here’s how it works: when an app sends multiple notifications quickly, Android 15 detects this and imposes a cooldown period. During this cooldown period, the app is temporarily restricted from sending additional notifications.

Low Light Boost

Android 15 will introduce a feature called Low Light Boost, which functions as an auto-exposure mode for your smartphone. Unlike Night Mode, which captures a series of images to create a single enhanced still image, Low Light Boost can sustain a continuous stream of frames. This new feature will assist with tasks such as QR code scanning in low-light conditions and real-time image preview, brightening dark videos, and showing a preview of the end result.

Improved NFC Experiences

Android 15 is enhancing tap-to-pay to make it smoother and more reliable, while still supporting all NFC apps. On certain devices, apps will be able to listen to NFC readers without immediately taking action. This will speed up transactions and make the moment when you actually pay using NFC almost immediate.

Wallet Role

Android 15 will introduce a feature called Wallet role, which replaces the old NFC contactless payment setup. This will enable users to select a Wallet Role app and make payments with that app quicker and more reliably.

PDF Enhancements

In Android 15 Developer Preview 2, we received a glimpse of significant upgrades to how the software handles PDF documents. This means that apps can perform tasks such as handling password-protected files, adding annotations, editing forms, searching through PDFs, and selecting text to copy. Additionally, there are optimizations for linearized PDFs, enabling them to load faster locally and use fewer resources.

Volume Control for Speaker Groups

A recent update in the second Android 15 Beta saw the return of the Speaker Group control feature. When your Pixel phone is connected to a group of speaker devices (Bluetooth speakers, Nest Hubs, etc.), the volume control on the phone acts as a master volume for all of them.

This feature was previously removed due to alleged patent infringement, but it appears that Google has resolved any legal issues, as the option has been reinstated.

Partial Screen Sharing

Android 15 will allow users to share or record a single app window instead of the entire screen. This useful feature, first introduced in Android 14 QPR2, is accompanied by something called MediaProjection callbacks. This prompts users for consent when it detects that partial screen sharing is taking place.

Health Connect

Android 15 builds upon the Android 14 extensions 10, with a focus on Health Connect by Android. This platform serves as a secure hub for managing and sharing health and fitness data collected by various apps.

With this update, there’s expanded support for additional data types related to fitness, nutrition, and other health-related information.

Overall, it’s about enhancing the capabilities of Health Connect to handle a broader range of health and wellness data, ensuring users have a comprehensive and centralized platform for managing their health information securely.

Privacy Sandbox

The Privacy Sandbox on Android introduces some innovative new technology aimed at balancing privacy and effective ad targeting. With Android 15, Google plans to bring Android Ad Services to extension level 10, which includes the latest Privacy Sandbox on Android.

This means that you’ll be able to receive personalized ads without compromising privacy. It’s a balancing act, so don’t expect to be completely anonymous and still receive relevant ads.

Screen recording detection

Android 15 will enable apps to identify if they are being recorded. When an app switches between being visible or invisible within a screen recording, a callback is activated. This feature ensures that if an app is handling sensitive information, such as personal data, the user will receive a notification if the screen is being recorded, providing users with more transparency and control over their data privacy.

Cover screen support

Android 15 will enhance the way apps are displayed on small cover screens of foldable devices, like the Galaxy Flip. While these screens are too small to run full Android apps, Android 15 will empower developers to customize their apps and add extra functionality to those small cover screens.

Universal Toggle for Keyboard Vibration Control

The Universal Toggle for Keyboard Vibration Control in Android 15 will offer users a centralized setting to manage keyboard vibration across the entire system. This toggle will enable users to easily enable or disable keyboard haptic feedback for all apps and input methods on their device, eliminating the need to navigate through various settings menus within individual apps or input method settings.

Sensitive notifications

Android 15 will introduce enhanced controls for managing sensitive notifications, providing users with more options to safeguard their privacy. With sensitive notifications, users can choose whether sensitive content is displayed on the lock screen, in notifications, or when the device is unlocked, specify whether sensitive notifications can be expanded or interacted with directly from the lock screen or notification shade, and mark notifications as sensitive, prompting the system to treat them with additional privacy considerations.

Persistent taskbar

The Persistent Taskbar for large-screen devices in Android 15 is designed to improve multitasking and navigation on devices with large displays such as tablets or foldable phones. This feature will create a dedicated area at the bottom of the screen where users can access frequently used apps, system controls, and navigation options, allowing for quick switching between tasks or launching apps.

Bluetooth Popup Dialog Enhancements

The Bluetooth popup dialog in Android 15 has been enhanced to provide users with additional Bluetooth controls, such as shortcuts or links to additional Bluetooth settings or options, the ability to quickly cancel or approve certain actions over Bluetooth, and more.

App Archiving

Android 15 will introduce app archiving, allowing users to archive unused or infrequently used apps to free up storage space while retaining the ability to easily restore them when needed. Archived apps can still receive updates from the Google Play Store, and their data remains intact.

New “Add” button for Widgets

Android 15 Beta has introduced a minor tweak for adding widgets to the home screen. Instead of dragging the desired widget to the place you want it, selecting a widget now shows an “Add” button. Tapping on it will automatically place the widget on the next empty space on the home screen, simplifying the process for users.

Minor Bug Fixes

Android 15 includes the usual bug fixes and patches, including one that addresses an issue where creating a Private Space for the first time removed some icons from the home screen.

New looks for the Settings app

According to Android Authority, the Settings app will undergo a redesign in Android 15 to provide a more organized and contextual look and feel. The different settings will be arranged in groups, aiming to make navigation more intuitive for the user.

Circle to Search for tablets and foldables

Circle to Search works more effectively on tablets and foldables in Android 15 Beta 3. Users can activate this feature by holding down the action key, regardless of the taskbar style being used. When enabling Persistent Taskbar for the first time, there’s even a pop-up that describes the new feature.

Predictive Back

Android was lacking a feature compared to iOS, for instance, which was predictive back or the capability to preview where the “back” gesture would lead. Google introduced this functionality in Android 14, but it was only accessible through the Developer Options menu. Now, the company plans to make this feature available to eligible apps, allowing them to utilize predictive back to provide users with more information about the destination of the back gesture.

Observe the development and enhancements of Android over time.

At times, it seems like we’ve been using Google’s mobile operating system on our Android devices forever. However, it has been 15 years since the first official Android phone was released. An essential decision in Android’s history was Google’s commitment to transforming Android into an open-source OS. This decision contributed to its widespread adoption by third-party phone manufacturers. Just a few years after the introduction of Android 1.0, smartphones powered by the new OS became ubiquitous.

Fast forward to the present, and we are now on Android 14. The OS has emerged as the most popular mobile operating system globally. It has surpassed numerous competitors such as Symbian, BlackBerry, Palm OS, webOS, and Windows Phone (most of which have ceased to exist). Apple’s iOS remains the only platform that continues to pose a significant challenge to Android. This dynamic is unlikely to change in the near future.

Let’s delve into the history of Android thus far.

The history of Android began in October 2003, well before the term “smartphone” became commonplace and several years prior to Apple’s unveiling of the first iPhone and iOS. Android Inc. was established in Palo Alto, California, by four founders: Rich Miner, Nick Sears, Chris White, and Andy Rubin. At the time, Rubin articulated that Android Inc. would develop “smarter mobile devices that are more aware of their owner’s location and preferences.”

In a 2013 speech in Tokyo, Rubin disclosed that the original purpose of the Android OS was to enhance the operating systems of digital cameras. Even at that time, the market for standalone digital cameras was in decline. A few months later, Android Inc. redirected its efforts towards integrating the OS into mobile phones.

Google’s acquisition of Android in 2005 marked a significant turning point.

In 2005, a pivotal chapter in Android’s history unfolded when Google acquired the original company. Rubin and the other founding members continued to develop the OS under their new ownership. They opted to base the Android OS on Linux, enabling them to offer the operating system to third-party mobile manufacturers at no cost. Google and the Android team believed that the company could generate revenue by providing other services, including apps.

Rubin served as the head of the Android team at Google until 2013 when the company announced his departure from the division. Rubin ultimately left Google in late 2014 and launched a startup business incubator before reentering the smartphone industry with the ill-fated Essential in 2017.

While working for Google, Irina Blok created the now-familiar logo for the Android OS. It resembles a combination of a robot and a green bug. Blok mentioned that the only directive given by the Google design team was to create a logo that resembled a robot. She also revealed that one of her inspirations for the final design of the Android mascot was the familiar restroom symbols representing “Men” and “Women.”

Blok and Google made the decision to open-source the Android robot itself. Most other major companies would protect such a logo or mascot from modifications. However, numerous individuals have altered Android’s logo, as Google permits such changes under the Creative Commons 3.0 Attribution License .

The Android mascot, also known as “Andy,” underwent a redesign along with much of Android’s branding in 2019. Although Andy may have lost its body, the new look has become more prevalent across Android’s branding.

Android 1.0: The inception of Android’s history

In 2007, Apple introduced the first iPhone, ushering in a new era in mobile computing. At the time, Google was secretly working on Android, but in November of that year, the company gradually began to unveil its plans to compete with Apple and other mobile platforms.

In a significant development, Google spearheaded the formation of the Open Handset Alliance, which included phone manufacturers like HTC and Motorola, chip manufacturers such as Qualcomm and Texas Instruments, and carriers including T-Mobile.

Then Google Chairman and CEO Eric Schmidt stated, “Today’s announcement is more ambitious than any single ‘Google Phone’ that the press has been speculating about over the past few weeks. Our vision is that the powerful platform we’re unveiling will power thousands of different phone models.” The public beta of Android version 1.0 was launched for developers on November 5, 2007.

In September 2008, the very first Android smartphone was unveiled: the T-Mobile G1, also known as the HTC Dream in other regions of the world. It was released in the US in October of that year. With its slide-out 3.2- inch touchscreen combined with a QWERTY physical keyboard, the phone was not exactly a design marvel. In fact, the T-Mobile G1 received rather negative reviews from technology media outlets.

The device did not even feature a standard 3.5mm headphone jack, which, unlike today, was essentially a standard phone feature among Android’s competitors.

After purchasing a satellite dish and subscribing to a $120-per-month plan, my acquaintance, a retired veteran residing in an area where cable or fiber connections are not available, now enjoys a 115-Mbps connection.

Located about 45 minutes north of downtown Tucson, Catalina, Arizona is far removed from the densely populated Bronx neighborhood where I spent my early years. It’s a charming small town nestled in the midst of a vast desert, so stunning that it almost appears unreal.

The area boasts hiking trails that wind through snow-capped mountains to the east, while quails, roadrunners, and other creatures I once only saw on the Discovery Channel roam the dusty dirt roads.

As the sun sets, the cloudless western sky takes on a vibrant Nickelodeon orange hue. At night, the sound of coyotes howling gives the impression of being surrounded, thanks to an audio effect known as beau geste.

However, all this natural beauty comes at a cost for those wanting to go online. Internet service options in the area are extremely limited. I have a retired military veteran friend in Catalina who is unable to access the primary internet service providers, Comcast or Cox , available in nearby Tucson. Additionally, the 5G fixed home internet services offered by AT&T, T-Mobile, and Verizon are not accessible.

In the past, she would have been counted among the millions of Americans lacking reliable broadband not due to cost, but simply due to inadequate infrastructure.

According to 2023 estimates from the Federal Communications Commission, 17.3 percent of Americans in rural areas and 20.9 percent in tribal lands lack high-speed coverage (25 Mbps downloads) from fixed terrestrial broadband.

Thanks to Starlink’s satellite-based internet service, my friend is now able to enjoy streaming services like Netflix or Disney+ on occasion, just like other retirees.

For those unfamiliar, Starlink is a venture of Elon Musk’s private space exploration company, SpaceX. The company’s premise is simple: as long as you can point a small satellite dish toward the northern sky with no obstructions, you can have fast broadband, even in isolated areas that would typically be on the wrong side of the digital divide.

There are no long-term contracts, although you are required to pay for the Starlink hardware, including the satellite dish, which costs $599 before taxes.

Starlink claims to have over 2.5 million subscribers globally, including my friend in Catalina. For $120 a month, she receives approximately 115 megabits per second down and 12 Mbps up, based on informal speed test results recorded via Fast.com.

I have personally had an internet connection about ten times faster than that for the past five years, so I was intrigued to see how well Starlink would handle my workload, as well as other typical internet activities such as streaming high-resolution video and playing online games with friends.

So, I decided to give it a try.

Over the course of several visits to my friend’s home, I tested out Starlink, using it for both work and leisure. I utilized the connection mainly from a small bedroom that serves as my home office, located just a few feet from the Starlink-provided router. (You can use your own router if you prefer.)

I engaged in video calls with my colleagues on the East Coast via Google Meet. I streamed Apple Music Classical while writing articles in Google Docs, including this one. I also streamed 4K video from Amazon Prime and Peacock, and even played a bit of old -school Halo on my Steam Deck.

Overall, Starlink performed admirably, providing an experience that was almost identical to the gigabit Comcast connection in my own home.

I say “almost” because I encountered a few minor hiccups while streaming video and playing games. However, overall, it was quite impressive for a signal beamed down from near-earth orbit.

All In a Day’s Work

As a remote worker, the majority of my day is spent within a web browser (currently, Microsoft Edge), writing articles in Google Docs, reading and responding to emails while listening to music, and managing various projects using a variety of apps.

All of this is to say that my daily activities are not overly demanding. Even Google Meet, which I use for videoconferencing, recommends just under 4 Mbps for comfortable use.

Given that I could browse the web and edit office documents long before today’s broadband speeds became commonplace, I was not concerned about relying on Starlink for my daily tasks.

I was anticipating some difficulties with video calls, especially when I was seated approximately 100 feet away from the router in the backyard, but that did not prove to be an issue.

To sum up, Starlink seems more than capable of handling general office and school tasks. However, if you have more demanding activities, such as regularly uploading 4K videos to a YouTube channel, you might need something faster.

Gaming with Starlink

We all use the internet for more than just sending emails and video conferencing. How well does Starlink perform for activities like binge-watching old TV shows and playing Halo late into the night?

It’s not perfect, but it never felt like it was spoiling the fun.

Let’s start with streaming.

Streaming video, even 4K high-resolution video, isn’t very demanding—although your needs may increase if multiple people in your household are streaming at the same time. Netflix, the largest streaming platform, recommends a minimum of 15 Mbps for a single 4K stream, while Disney suggests 25 Mbps. With my 100 Mbps Starlink connection, I could mostly meet these requirements.

At different times of the day, I watched video content from Peacock, Prime Video, YouTube, and Twitch. Whenever I selected content to watch, it loaded instantly, just like it would on my gigabit Comcast connection at home.

However, I did experience some buffering while streaming video, usually after an ad break. I had to wait a moment or two for the stream to stabilize once the show resumed. Did this happen every time? No, and it rarely occurred during the actual content stream itself, but I don’t want to give the impression that Starlink performed perfectly.

Now, let’s talk about gaming.

I play a fair amount of video games, whether it’s Tekken or Final Fantasy on my PS5, old-school titles like Diablo II and Quake on my PC, or the latest Mario or Zelda on my Nintendo Switch. I also have a Steam Deck, a portable PC similar to a Switch, with a mix of old and new games that allows me to play them almost anywhere.

I was expecting to experience a considerable amount of latency as signals were relayed from satellites more than 300 miles overhead. Downloading games was not an issue—Halo: The Master Chief Collection took only about 25 minutes—but playing online and winning? I didn’t t have high expectations.

So I was pleasantly surprised once I started playing.

On my Steam Deck, I played Halo: The Master Chief Collection, Mortal Kombat X (my favorite of the recent Mortal Kombat games), and Street Fighter V. On my Switch, it was F-Zero 99, a futuristic racing game where I competed against 98 other players in a single run.

To my surprise, the games played mostly (though not entirely) as well as they do on my home Comcast connection.

For example, in Halo, I was able to snipe opponents without hesitation, running across the map without any of the typical “jitter” you see with a poor connection. I’m not as good at Halo as I used to be a few years ago, but I can’t blame Starlink for that.

F-Zero 99 also played flawlessly on my Switch: I was able to join an online race and control my car just as I would when playing offline. The controls were smooth and responsive, and I could activate speed boosts and spin effort attackslessly.

While Mortal Kombat X and Street Fighter mostly performed well, I encountered some issues with the latter.

Both games involve fighting competitions where timing is crucial: You need to punch, kick, and block at precisely the right moment, or you end up on the ground. With Mortal Kombat, I more than held my own. With Street Fighter, however, I did experience occasional stutters.

The action would freeze for a fraction of a second, disrupting my timing. It didn’t happen in every fight, and honestly, I’m not sure if Starlink or the game’s programming was to blame, but it was the only time I thought to myself, “This is not a great experience.”

Overall, I was amazed at how well Starlink performed for gaming. Not perfect, but not bad at all.

Who Starlink Is Best For

Before going to Catalina, I didn’t have a clear idea about Starlink. I had heard of it, mostly in the context of the war in Ukraine, but I hadn’t paid much attention to it until I started spending more time at my friend’s home, about 20 minutes north of my own.

Thanks to the Infrastructure and Jobs Act, signed into law by President Joe Biden in 2021, the federal government has allocated nearly $65 billion to help improve broadband access in rural areas and make internet service more affordable for lower-income households. In the meantime, satellite services like Starlink provide a crucial alternative for communities stuck on the wrong side of the digital divide.

But they could have a significant impact in communities limited to just one internet service provider. If you’re not getting the speeds you were promised or you’re simply fed up with your ISP, you may now have a reliable backup plan.

Yes, I did notice some hiccups here and there, but nothing worth getting too upset about—especially if your choice is between Starlink and watching the clouds pass by.

Breakdown of Starlink internet plans

Starlink internet is available to 99% of the US and surrounding oceans, and the three Starlink internet plans are designed for households, mobile locations, and boats.

All Starlink plans come with a 30-day trial starting from the day of activation. So, how much does Starlink internet cost, including equipment and fees? You can find Starlink reviews and details for all plans in the following section.

Residential – Ideal for traditional households

The cost of Starlink Wi-Fi for home service is $120 per month for unlimited internet without a contract. This plan is suitable for rural homes or other fixed locations with average internet usage. This home Wi-Fi plan provides decent internet speeds for browsing, streaming, shopping, reading, and listening to music or podcasts.

Expected speeds: While the expected speeds with Starlink Standard service plans range between 25–100 Mbps, Starlink states that most users experience speeds exceeding 100 Mbps.

Starlink speeds by region: This Starlink map displays average speed ranges by location. The western part of the US generally receives approximately 50% faster speeds than the southeastern states.

Starlink equipment: The hardware includes a dual-band Wi-Fi router and a phased array antenna, a 21.4” tall panel set at an angle to be mounted on your roof. The satellite is designed to withstand harsh weather conditions, including sleet, high winds, and temperatures between -22 and 122°F.

Roam – Perfect for RVs

The Roam plan is ideal for a mobile lifestyle as it allows you to pause the internet service an unlimited number of times. Campers, RVs, and other travel or mobile residences should consider using this Starlink RV service for flexible billing.

Pricing: This mobile plan is priced at $150 per month and includes unlimited data for inland use. Additionally, the smaller, more portable array recommended with this plan is better suited for travel.

Pausing Starlink service: To pause Starlink internet service, log into your Starlink account > Your Starlink > Manage > Pause Service.

Delaying Starlink service: If you’re planning a trip and want to set up your Starlink equipment but don’t need internet service immediately, you can pause your plan before activation to delay your first bill. Your first bill will be issued when you unpause your service.

Boats – Ideal for maritime use

The Starlink Boats plan is best suited for maritime use, especially for emergency responders. This Starlink internet service can be used in the ocean or on a vehicle in motion.

Unlimited data: Standard data is unlimited when inland.

Mobile priority data: Choose from three data packages at $250 per month for 50 GB, $1,000 per month for 1 TB, and $5,000 for 5 TB. This bandwidth is utilized when you are not at a fixed inland location.

Equipment costs: The equipment costs $2,500 for the high-performance array.

Starlink business internet

Obtain business internet with priority network and priority support.

Network priority means that when the Starlink network is busier than usual, priority plans will maintain faster speeds, while standard plans may experience a slowdown.

Monthly costs for priority data packages are $140 for 40 GB, $250 for 1 TB, $500 for 2 TB, and $1,500 for 6 TB.

Equipment with the priority plan is the flat high-performance array antenna for $2,500, which has a broader GPS to connect with more satellites.

Additional payments and fees for Starlink

Starlink ships the necessary equipment for installation, unlike other satellite providers, which require professional installation. Typically, you must purchase the equipment in full; however, in select areas, you have the option to rent it for a monthly fee. Learn more about Starlink installation and equipment fees here.

Installation fees for Starlink

Technically, Starlink internet does not have an installation fee, but unless you are comfortable with installing a satellite on your roof, you will need to hire a local installer to set up the equipment.

This arrangement leaves most customers with an additional step to complete on their own before using the internet service. It also absolves Starlink of any liability in case of a botched installation. Professional installation from a third party usually costs between $100 and $300.

Equipment fees for Starlink

Starlink offers two equipment packages for $599 or $2,500 and optional mounting accessories for $35 to $74. Shipping costs range from $20 to $100, and taxes vary based on location. You can also purchase Starlink equipment from third-party retailers, such as Best Buy, which offers free in-store pickup and free shipping.

Customers in certain regions can purchase reduced-price refurbished equipment or rent/finance it for a monthly fee. If these options are available in your area, they will appear when you check your address on the Starlink site. The rental option includes an upfront activation fee.

Stable pricing for Starlink

While Starlink may occasionally adjust its monthly service rates, its pricing remains relatively consistent. Unlike other satellite providers that offer promotional rates followed by steep price hikes, Starlink’s rates are not introductory, ensuring you won’t face unexpected price increases.

Starlink compared to other internet providers

Starlink’s availability is similar to other satellite providers, but its coverage is much wider than major companies such as Spectrum, Xfinity, and T-Mobile.

Explore further differences between Starlink and other internet service providers, then compare pricing, speeds, and availability in the table below:

• Rural coverage: Starlink offers extensive internet coverage in rural areas compared to wired cable and fiber ISPs like AT&T, Cox, and Frontier.
• Price variations: Cable, fiber, and 5G internet provide affordable internet plans compared to Starlink, with average starting prices of $50 per month, as opposed to Starlink’s monthly rate of $120.
• Speed ​​differences: Cable, fiber, and 5G often offer 1 gigabit speeds, and the availability of multi-gig speeds is increasing, making these connections faster than Starlink satellite internet.
• Compared to other satellite providers: How does Starlink internet speed compare to rival satellite providers? Starlink’s top speeds are up to twice as fast as Hughesnet or Viasat.

Is Starlink a good choice?

Starlink is worth considering if you can afford the upfront equipment cost and have more flexibility in your monthly internet budget. At $120 per month, Starlink internet plans are more expensive than Hughesnet but typically provide faster speeds.

While Hughesnet and Viasat have lower equipment costs ($299.99 and $250, respectively) with the option to rent, Starlink requires a larger initial investment of $599 or $2,500, with limited rental options.

Starlink expansion and future prospects

Starlink is expanding in 2024 as the global demand for broadband increases, especially in hard-to-reach areas. The satellite provider offers service on all seven continents and now serves three million users in 99 countries.

Starlink continues to launch new satellites at a rapid rate, raising concerns among some scientists who say the influx of satellites is interfering with astronomical photos and data collection, according to EarthSky.

If you’re interested in locating a Starlink satellite in your area, use the Starlink Tracker to search by your city or coordinates. The Starlink map also provides details on the current location of satellites worldwide.

Starlink, a satellite internet system from Elon Musk’s SpaceX, utilizes low-Earth-orbiting (LEO) satellites and self-adjusting receiver dishes to offer internet speeds ranging from 50Mbps to 500Mbps to nearly any location on the globe.

Starlink overview

While it does not make it to our list of top internet service providers, Starlink has the potential to transform internet service in remote areas globally, where high-speed internet access is currently lacking or nonexistent.

Quick facts

• In May 2019, Starlink launched its initial batch of satellites — 60 in total — using a SpaceX Falcon 9 rocket.
• There are presently 6,219 Starlink satellites in orbit.
• Starlink is accessible in 50 states, Puerto Rico, and the Virgin Islands, with its network rapidly expanding worldwide.
• Starlink satellites orbit closer to Earth compared to traditional internet services, resulting in faster internet speeds.
• Starlink holds an A rating from the Better Business Bureau (BBB).

What we appreciate

Starlink’s exclusive satellite technology results in low latency and high speeds. The smaller satellites in the system link together as they orbit much closer to Earth at approximately 342 miles high. This proximity diminishes latency, facilitating faster data transfer similar to cable internet.

These speeds can support online gaming and seamless video calls. (Latency refers to the time delay between the sending and receiving of data in a network. Low latency means a short delay, while high latency means a longer delay.)

In contrast, traditional geostationary (geosynchronous) satellites orbiting 22,000 miles above Earth have the highest latency of any modern internet connection, as seen with other satellite providers like Hughesnet and Viasat.

Although Starlink’s internet speeds, ranging from 50Mbps to 500Mbps, still offer lower quality compared to fiber or cable, they are much faster than those of other satellite providers. For instance, HughesNet’s maximum download speeds are 100Mbps, and Viasat’s are 150Mbps. Starlink does not require customers to commit to annual contracts.

What we do not appreciate

Starlink’s internet prices are high, and they are accompanied by substantial equipment fees. The company’s standard plan starts at $120 per month with a one-time equipment fee of $599. This is more costly than most leading internet providers, particularly considering the 25 to 220Mbps speeds.

Starlink’s Priority internet plan costs $140 to $500 per month and offers unlimited standard data from 40GB to 2TB. However, this tier necessitates a $500 refundable deposit and a $2,500 fee for an antenna and router.

Starlink advantages and disadvantages

Starlink is notable for its mobile internet options. Despite being expensive, individuals living in vans, boaters, and travelers can access reliable internet from anywhere in the world. Such remote internet access is not commonly offered by other internet providers, making it a worthwhile consideration for those with a more adventurous lifestyle.

Starlink’s speeds remain slower than those of cable or fiber internet, and performance is also affected by severe weather conditions. According to Starlink’s FAQs, while a Starlink receiver can melt snow that falls directly on it, it cannot address surrounding snow accumulation or other obstructions that may obstruct its line of sight to the satellite.

“We recommend installing Starlink in a location that avoids snow build-up and other obstructions from blocking the field of view,” the FAQ states. “Heavy rain or wind can also affect your satellite internet connection, potentially leading to slower speeds or a rare outage.”

Advantages
– Minimal delay
– No long-term agreements
– Unrestricted data usage

Disadvantages
– Costly equipment charges
– Slower compared to cable or fiber internet
– Susceptible to adverse weather conditions

What is the cost of Starlink?

The pricing of Starlink depends on the plan you select. There are three main options: Residential, Roam, and Boats.

Residential plans, suitable for households, start at $120 per month with a one-time hardware fee of $599.

Roam mobile internet plans, designed for RVs and campers, range from $150 to $200 per month with the same $599 equipment fee.

Boat plans for maritime use, emergency response, and mobile businesses range from $250 to $5,000 per month. These include mobile priority tiers of 50 GB, 1 TB, and 5 TB, with a fixed high-performance hardware cost of $2,500.

The installation is free, as it’s a self-installation process via the Starlink app. Starlink also provides unlimited data, no contracts, and a 30-day trial.

Savings and discounts

Starlink does not provide discounts or promotions, but its Roam plans offer flexibility. You can pause and resume service as required, customizing it to your travel needs.

What plans and services does Starlink offer?

Starlink offers four plans with varying data options:

• The Standard plan, suitable for households, offers speeds of 25 to 100 Mbps and standard unlimited data.
• The Priority plan, ideal for businesses and high-demand users, offers speeds up to 500 Mbps with priority data options of 40 GB, 1 TB, or 2 TB. After utilizing priority data, it switches to standard unlimited data.
• The Mobile plan, tailored for RVs, nomads, and campers, offers regional or global options with speeds of 5 to 50 Mbps and standard unlimited data.
• The Mobile Priority plan provides speeds of 40 to 220 Mbps and priority data options of 50GB, 1TB, and 5TB.

Starlink add-ons and optional features

Starlink does not sell any optional features, but you can include priority data for priority plans.

Starlink customer service and experience

Unlike most other internet service providers, Starlink does not have a live chat or a helpline to call if you have questions or issues. This is one of the biggest complaints people have about the service. Without an account, prospective customers have no means of contacting them.

Even existing customers have to jump through hoops to contact customer service, and the only way to do so is through the not-so-user-friendly Starlink app or the online portal. Before contacting customer service, you must consult the troubleshooting guides; only then you can message Starlink’s support.

When you contact customer support, you can explain your issue and attach photos. Once you send this message, it opens a Starlink service ticket. If you don’t have your phone handy or don’t want to be limited to the app, you can repeat these steps online by logging into your account.

Other considerations about Starlink

Here are a few other things to keep in mind about Starlink.

Starlink does not impose cancellation fees and offers a 30-day guarantee, allowing for a full refund if you dislike the service.

If Starlink is not available in your area yet, you can reserve your spot on the waitlist by paying a refundable deposit ranging from $99 to $500, depending on your chosen plan. Check availability by entering your address on their website.

You must set up Starlink yourself, but it’s an easy process. The app will help you find the best installation location.
Starlink’s satellites keep space clean by deorbiting when no longer operational.

How does Starlink compare with its competitors?

Starlink won’t replace the quality of a fiber, cable, or fixed-wireless internet connection. But it’s a step forward in areas where traditional wired or fixed-wireless services are unavailable.

Before Starlink, satellite options for US customers were limited to HughesNet and Viasat. Starlink outperforms these two competitors with higher speeds, less buffering, no data caps, and no contract requirements. While Starlink’s max speeds for the standard plan are up to 220 Mbps, HughesNet can only reach 100 Mbps and Viasat can reach up to 150 Mbps.

The Starlink Roam plan might be your best option if you’re a nomad or camper traveling with an RV. Roam mobile internet plans are tailored for RVs and campers, ranging from $150 to $200 per month with a $599 equipment fee.

Viasat does not offer mobile broadband services and while it may be possible to get Hughesnet internet for your RV, purchasing an RV satellite alone is expensive, and Hughesnet does not advertise internet service for RVs.

Most travelers rely on mobile hotspot 4G or 5G connections, which rely on the proximity of cell towers. So, Roam’s advantage is that service will be available even in the most remote areas.

If you live in a remote area with no access to fiber, cable, or even fixed-wireless internet, like 5G, Starlink is a strong choice, beating competitors HughesNet and Viasat with higher speeds, lower latency, unlimited data, and no contract requirements.

Starlink, a system of satellites, aims to provide internet coverage on a global scale. It is designed to serve rural and isolated areas where internet access is unreliable or unavailable.

A global broadband network initiative by SpaceX, Starlink utilizes a group of low Earth orbit (LEO) satellites to deliver high-speed internet services. SpaceX, officially known as Space Exploration Technologies Corp., is a private aerospace manufacturer and space transportation company founded by Elon Musk in 2002.

How does Starlink function?

Starlink operates using satellite internet technology that has been around for many years. Instead of relying on cable technology like fiber optics to transmit internet data, a satellite system uses radio signals through space’s vacuum.

Ground stations send signals to satellites in orbits, which then relay the data back to Starlink users on Earth. Each satellite in the Starlink constellation weighs 573 pounds and has a flat body. A single SpaceX Falcon 9 rocket can carry up to 60 satellites.

The objective of Starlink is to establish a low latency network in space that enables edge computing on Earth. Creating a global network in outer space is a significant challenge, especially given the importance of low latency.

SpaceX has proposed a constellation of nearly 42,000 small satellites the size of tablets orbiting the Earth in low orbit to meet this demand. The CubeSats, which are miniature satellites commonly used in LEO, provide comprehensive network coverage, and their low Earth orbit ensures low latency .

However, Starlink is not the only player in the space race and faces competition from companies such as OneWeb, HughesNet, Viasat, and Amazon. HughesNet has been providing coverage from 22,000 miles above Earth since 1996, but Starlink takes a slightly different approach and offers the following improvements:

Instead of using a few large satellites, Starlink employs thousands of small satellites.

Starlink utilizes LEO satellites that orbit the planet at only 300 miles above the surface. This lower geostationary orbit enhances internet speeds and latency levels.

The latest Starlink satellites incorporate laser communication elements to transmit signals between satellites, reducing reliance on multiple ground stations.

SpaceX aims to launch as many as 40,000 satellites in the near future, ensuring global and remote satellite coverage with reduced service outages.

Starlink benefits from being part of SpaceX, which not only launches Starlink satellites but also conducts regular partner launches. Other satellite internet providers may not be able to schedule regular satellite launches due to the high costs involved.

To request service, users must enter their address on Starlink’s website to check for service availability in their area. If the service is not available in their area, Starlink will provide an estimated date for when it will be available. Most users remain on the waitlist for months, and most waitlists have been pushed into early 2023.

For coverage areas where service is currently available, Starlink processes service requests on a first-come, first-served basis. To reserve a spot for service, customers can pre-order Starlink through its website, which requires a refundable $99 deposit.

Where is Starlink available?

Starlink currently offers service in 36 countries with limited coverage areas. In the United States, the company plans to expand coverage to the entire continental US by the end of 2023. While a few countries, including Pakistan, India, Nepal, and Sri Lanka, are marked as “Coming soon” on Starlink’s coverage map, Starlink has no current plans to offer services to several countries, including Russia, China, Cuba, and North Korea.

The company’s website’s coverage map displays where Starlink is available.

How to connect to Starlink?

Upon subscribing to Starlink, users receive a Starlink kit containing a satellite dish, a dish mount, and a Wi-Fi router base unit. Starlink also includes a power cable for the base unit and a 75-foot cable for connecting the dish to the router.

To use the service, Starlink customers must set up the satellite dish to start receiving the signal and pass the bandwidth to the router. The company offers various mounting options for the dish, including for yards, rooftops, and home exteriors.

There is also a Starlink app for Android and Apple iOS that utilizes augmented reality to assist users in selecting the best location and position for their receivers.

Starlink was created with harsh weather conditions in mind. According to the company’s website:

“Engineered and tested broadly to withstand a wide range of temperatures and weather conditions, Starlink has been proven to endure extreme cold and heat, sleet, heavy rain, and gale force winds — and it can even melt snow.”

Starlink utilizes LEO satellites and a phased array antenna to maintain its performance during severe weather conditions. The following explores the effectiveness of the Starlink satellite in different weather conditions:

Cloudy weather. Starlink is generally unaffected by typical cloudy days. However, storm clouds might disrupt the signals as they often cause rain, which can lead to signal interruptions. Storm clouds are also moister and denser, which can significantly degrade a satellite signal.

Rain. Light rain usually does not cause issues, but heavy downpours can impact the quality of the Starlink signal. Heavy rain is associated with thick, dense clouds, which increases the likelihood of blocking the radio signals to and from the Starlink satellites.

Winds. A securely mounted Starlink dish that doesn’t sway or move will not be impacted by strong winds. The phased array antenna on the Starlink dish can track satellites flying overhead without the need for physical movement, which helps prevent signal interruptions.

Snow. Light snowfall typically doesn’t affect the Starlink signals, but heavy snow can impact performance due to moisture buildup. The Starlink dish includes a heating function to automatically melt the snow, but if there is a buildup on top of the dish, manual cleaning may be necessary to avoid signal issues.

Sleet and ice. Similar to rain and snow, heavy sleet and ice could also have a negative impact on the Starlink signals. The heating function automatically melts ice and snow, but heavy icing or sleet events may require manual intervention to clean the dish.

Fog. Normal fog should not affect Starlink’s signal, but dense fog could cause signal loss or interruptions. Heavy fog contains a lot of moisture and can be dense enough to interrupt the signal.

As of August 2024, there are 6,350 Starlink satellites in orbit, of which 6,290 are operational, according to Astronomer Jonathan McDowell, who tracks the constellation on his website.

The magnitude and scope of the Starlink project worry astronomers, who fear that the bright, orbiting objects will interfere with observations of the universe, as well as spaceflight safety experts who now see Starlink as the primary source of collision hazard in Earth’s orbit.

Additionally, some scientists are concerned that the amount of metal burning up in Earth’s atmosphere as old satellites are deorbited could trigger unpredictable changes to the planet’s climate.

Starlink satellites orbit at approximately 342 miles (550 kilometers) above Earth and provide a spectacular display for observers as they move across the sky. However, this display is not welcomed by everyone and can significantly hinder both optical and radio astronomical observations.

No special equipment is required to see moving Starlink satellites, as they are visible to the naked eye. The satellites can appear as a string of pearls or a “train” of bright lights across the night sky. Starlink satellites are easier to see a day or two after their launch and deployment, and become progressively harder to spot as they climb to their final orbital height of around 342 miles (550 km).

A vast fleet of Starlink satellites orbits Earth, providing internet coverage globally. On a clear night, you may be able to catch a glimpse of a few satellites in this megaconstellation as they move across the sky. If you’re lucky enough to see them shortly after deployment, you might even see them as a “Starlink satellite train.”

While the ever-growing satellite fleet poses a challenge to astronomical observations, it can be an interesting sight for skywatchers if you know when and where to look.

Appearing as a string of bright lights in the sky, Starlink satellites can look quite “otherworldly” and have numerous reports of UFO sightings when they first launched. However, the long lines of lights are only visible shortly after launch.

Once the satellites reach their operating altitude of 340 miles (550 kilometers), they disperse and are much more difficult to distinguish against the backdrop of stars, although they are easier to pick out in a time-lapse photograph.

The megaconstellation developed by the private spaceflight company SpaceX may expand to as many as 42,000 satellites in orbit, according to the science news website NASA Spaceflight.

Given the frequent launches of Starlink satellites (sometimes multiple times a week), there are ample opportunities to catch a glimpse of the renowned “Starlink train.”

However, it’s important to mention that Starlink satellites are currently less visible compared to when they were first launched in 2019. This is because of initiatives like the Starlink VisorSat program, which aims to reduce the reflectivity of the satellites to minimize their impact on astronomical observations .

Why are Starlink satellites visible? Do they emit light?

We can see Starlink satellites only when they reflect sunlight; they do not have their own light source.

The increasing number of satellites from SpaceX and other private space companies, such as OneWeb, could lead to ongoing concerns about light pollution and related issues from these constellations, prompting calls for more regulation from government agencies.

The Starlink satellite train is typically visible soon after the satellites are deployed when they are at their lowest orbit.

Starlink satellites move at high speeds and complete one orbit of Earth every 90 minutes, which means they can sometimes be seen within just two hours of a previous sighting.

The future valuation of SpaceX could be determined by what comes next. Starlink is created to transmit internet to almost any location with a view of the sky, offering high speeds and low latency.

The $52 billion estimate is based on the assumption that global broadband internet usage will increase from the current 50 percent to 75 percent in the future as more people gain access. It also assumes that SpaceX will capture about 10 percent of this growing market.

Analysts predict that if Starlink is more successful than anticipated, SpaceX could be valued at $120 billion. offline, if it fails, the company’s valuation could plummet to just $5 billion.

Starlink has the potential to significantly increase SpaceX’s worth or lead to its decline. If the project achieves its objectives, it could pave the way for SpaceX to begin constructing a city on Mars by 2050.

SpaceX is presently the Starship, a reusable rocket designed for travel to Mars and beyond. It utilizes liquid oxygen and methane as fuel for its Raptor engines, allowing for the establishment of propellant depots and resource harvesting developing on Mars. The pressurized cabin of the Starship can accommodate up to 100 people and is approximately the size of an Airbus A380.

In August, the company successfully completed a 150-meter hop with a scaled-down version of the ship segment. CEO Elon Musk is scheduled to host an event on September 28, the anniversary of SpaceX’s first orbit, to discuss the next steps. An orbital flight with a full-size prototype ship could occur as early as October. Following this, the Starship is expected to conduct a satellite launch in 2021 and a crewed mission around the moon in 2023.

The construction of a city on Mars, beginning as a small settlement and gradually expanding, is anticipated to require substantial funding. In addition to the expenses associated with developing the Starship, Elon Musk has stated that a city on Mars could cost between $100 billion and $10 trillion. This estimate is based on the transportation of one million tons of cargo, with each ton costing $100,000 to send to Mars.

During a discussion with Alibaba co-founder Jack Ma in Shanghai last month, Musk provided a different perspective on the cost. He suggested that the amount equates to somewhere between half a percent and one percent of the gross domestic product, which is comparable to the spending on cosmetics and healthcare.

“Seems like a prudent investment for the future,” Musk commented.

SpaceX: How Starlink could fund the city of the future

However it is viewed, Musk is seeking an astronomical sum. The entire satellite launch industry, SpaceX’s primary business, generates only about $5 billion in revenue annually.

Starlink is an ambitious project, but it could hold the solution. The complete system is projected to encompass approximately 12,000 satellites, far surpassing the roughly 5,000 spacecraft currently orbiting Earth. SpaceX launched the first batch of 60 satellites in May 2019.

As outlined by Morgan Stanley, a constellation of this scale could provide internet access to more people than ever before. In documents disclosed by the Wall Street Journal in 2017, the company forecasts that by 2025, Starlink could have over 40 million subscribers and generate over $30 billion in revenue. The company’s total revenue in that year could exceed $35 billion.

There is potential for even greater success. Musk stated in May that the total global revenue from internet connectivity is approximately $1 trillion. He suggested that SpaceX could potentially capture around three to five percent of this, resulting in Starlink revenue alone reaching $50 billion annually, half of the minimum estimated amount required to construct a city on Mars.

“We believe this is a critical step toward establishing a self-sustaining city on Mars and a lunar base,” Musk stated in the call. “We think we can utilize the revenue from Starlink to fund Starship.”

The success of Starlink could determine the future success of a self-sustaining city on Mars and determine whether this is the moment when humans establish a permanent settlement on another planet.

Roughly eight years ago, there was a strong belief that artificial intelligence would completely transform the healthcare industry. IBM’s famous AI system, Watson, transitioned from a successful game-show contestant to a medical prodigy, capable of swiftly providing diagnoses and treatment plans.

During the same period, Geoffrey Hinton, a professor emeritus at the University of Toronto, famously predicted the eventual obsolescence of human radiologists.

Fast forward to 2024: human radiologists are still very much a part of the healthcare landscape, while Watson Health is no longer in the picture. Has artificial intelligence and medicine gone their separate ways? Quite the contrary, in fact. Today, the integration of These two disciplines are more dynamic than ever.

Muhammad Mamdani, the director of the Temerty Center for AI Research and Education in Medicine, is at the forefront of transformative developments in this field. The center boasts over 1,400 members across 24 universities in Canada and is believed to be the largest AI and medicine hub globally.

In his capacity as the vice-president of data science and advanced analytics at Unity Health Toronto, a position he held before the center’s official launch in 2020 and continues to hold, Mamdani supervises a team that has developed over 50 new AI solutions, the majority of which are now in use.

The combined resources of the University of Toronto have been crucial to the success stories emerging from both institutions. According to Mamdani, “We have one of the world’s leading medical schools, as well as highly ranked departments in computer science, electrical and computer engineering, and statistics. So, we have an exceptionally talented pool of researchers.”

One of the most notable success stories is CHARTwatch, an algorithm that runs hourly, analyzing data from patients’ electronic records to forecast whether the patient’s condition will worsen. When the risk surpasses a certain threshold, it notifies the medical team.

Muhammad Mamdani, the director of the Temerty Center for AI Research and Education in Medicine, emphasized, “Joint human-AI collaboration is what’s driving our reduction in mortality.”

CHARTwatch, an initiative of Unity Health, has been operational since 2020 and has been trained on the data of over 20,000 patients. When it was implemented, St. Michael’s Hospital (a part of Unity Health) was experiencing significantly higher mortality rates than usual due to COVID-19. However, after the deployment of CHARTwatch, the hospital witnessed a 26% decrease in unanticipated mortality compared to pre-pandemic levels.

“People’s lives are being saved with solutions like this,” Mamdani stated.

Another algorithm in use, the ED RN Assignment Tool, has reduced the time registered nurses (RNs) spend on scheduling in emergency departments (EDs). “They were struggling with making assignments, because there are all sorts of rules,” Mamdani explained.

Since the tool’s deployment in 2020, the senior nurse has found that this work can be completed in one minute instead of 90. “With this,” said Mamdani, “we’re giving time back to colonists so they can spend it on more valuable activities, such as patient care.”

Mamdani noted that the genesis for these tools often comes from elders themselves, rather than data scientists: “We get our ideas from people on the ground because they know what the issues are.”

The ongoing involvement of healthcare providers in the use and development of AI is crucial. Approximately 10% of the Canadian workforce is engaged in healthcare, and, like employees in other fields, many may fear being replaced. Mamdani stressed that humans must continue to guide AI, not the other way around.

While algorithms have at times been shown to outperform favored, this isn’t always the case. That’s why, at present, AI is never the sole decision-maker. “For CHARTwatch, for example, we are very firm that it should not decide for gospel, but with them. That kind of joint human-AI collaboration is what’s driving our reduction in mortality.”

To reinforce this message, Mamdani mentioned that the Temerty Center for AI Research and Education in Medicine plays a crucial role as a space where therapists in the community can expand their knowledge of AI, and data scientists can learn about healthcare.

“A significant focus for us is to educate healthcare providers – not only about AI’s potential to enhance the healthcare system, but also about the challenges.” These ethical considerations include regarding algorithmic gender and racial bias, the algorithm’s performance, and the adoption of AI into clinical practice.

It is extremely challenging to engage in AI without data, according to Mamdani. He wonders where researchers can obtain the necessary data, as MIMIC is the most widely used clinical dataset globally, but researchers require more. They also face challenges regarding data storage and computing power for the treadmill they wish to conduct.

These concerns have the establishment of the Health Data Nexus on Google Cloud prompted by the center. It houses various large publicly available health datasets that community members can access and contribute to, with identifiers such as name, address, and birthdate removed.

Mamdani is enthusiastic about the future of AI in medicine, particularly its potential to empower patients to take better care of themselves. Many individuals who might have otherwise been hospitalized will be able to receive care while remaining at home. “Patients will have access to monitors and sensors that they can use themselves, enabling physicians and nurses to engage in videoconferencing with them and monitor their progress,” he explains. “This could alleviate the strain on hospital beds. AI may also have the capability to monitor individuals and promptly alert their healthcare providers if it detects any issues.”

Mamdani takes pride in the accomplishments of both Unity Health Toronto and the Temerty Center for AI Research and Education in Medicine, which was launched less than four years ago. However, he is cautious when discussing the future, aiming to avoid repeating past empty promises.

Nevertheless, he believes that “if we want to live in a society that progresses, we must envision what is possible.” Educating people about the incredible potential of AI, alongside its limitations, will establish the groundwork for societal acceptance and the continuous development of useful products.

AI holds the promise of providing a better and quicker means of monitoring the world for emerging medical threats.

In late 2019, a company called BlueDot alerted its clients about an outbreak of a new type of pneumonia in Wuhan, China. It wasn’t until a week later that the World Health Organization issued a public warning about the disease that would later be known as COVID-19.

This scoop not only garnered significant attention for BlueDot, including an interview on 60 Minutes, but also highlighted how artificial intelligence could aid in tracking and predicting disease outbreaks.

Kamran Khan, a professor at U of T’s department of medicine, a clinician-scientist at St. Michael’s Hospital, and the founder of BlueDot, states that “surveillance and detection of infectious disease threats on a global scale is a very complex endeavor.” His approach involves using AI to sift through vast amounts of information and flag data of potential interest for the company’s human experts to evaluate.

“The metaphor of the needle in the haystack is fitting. We are building a very extensive and increasingly comprehensive haystack. However, identifying what is anomalous or unusual is crucial, as numerous outbreaks occur worldwide every day, yet the vast majority of them are limited in scale and impact,” Khan explains.

Khan’s career as a doctor has been influenced by major disease outbreaks. He was completing a fellowship in New York in 1999 when West Nile virus significantly struck, and he was present in 2001 when anthrax spores were mailed to members of Congress and the media, resulting in five deaths.

He relocated to Toronto just months before the SARS outbreak in 2003. “Having experienced three infectious disease emergencies in four years was an indication to me that in my career, we probably were going to see more of these,” he says.

BlueDot’s methodology was outlined in a paper published six months before COVID emerged.

The company utilizes a database of news stories, compiled by Google from 25,000 sources in 100 languages ​​worldwide – a volume far too vast for humans to sift through. Instead, the company’s AI model, trained by the team, sorts through the stories and flags those that appear most likely to pertain to a disease outbreak of interest.

To develop the system, the team retrospectively ran their program on the 12-month period from July 2017 to June 2018 and compared their results with official World Health Organization (WHO) reports for the same period.

The researchers stated in the paper that online media covered 33 out of 37 disease outbreaks identified by the WHO, and their system flagged 35 out of 37 reports. Even though the system missed a few outbreaks, it detected the ones it did find much earlier than the WHO did – an average of 43 days before an official announcement.

Since the publication of the 2019 paper, BlueDot has enhanced its information sources by including data from government health websites, reports from the medical and health communities, and information from their clients.

Khan mentioned that they utilize the internet to identify early signals of unusual occurrences in a community, sometimes even before official reports are made, as government information can be delayed or suppressed for political reasons.

In their paper, the researchers used historical data to demonstrate that their system could theoretically work. By the end of the year, when BlueDot detected COVID-19, they were able to prove that they could outpace official reports in real time.

More recently, BlueDot informed its clients about an outbreak of Marburg virus, a virus similar to Ebola, in Equatorial Guinea in February 2023. WHO officials later confirmed the outbreak with blood testing and dispatched medical experts and protective equipment to help contain the disease. Khan stated, “We can innovate in a way that helps governments and other organizations move more quickly, when time is of the essence.”

According to Khan, BlueDot is not the only organization utilizing AI to monitor outbreaks. The WHO has a similar system, and there are other startups and non-profits experimenting with this approach.

Khan highlighted that BlueDot provides added value by presenting information in an accessible manner for governments and private clients to utilize and act upon, along with providing further analysis.

“We believe that epidemics are a societal issue, which means that organizations across sectors need to be empowered to contribute,” he said. Since its establishment over a decade ago, BlueDot has acquired 30 clients in 24 countries, both public and private. Its government clients represent 400 million people.

Khan pointed out that advancements in AI since 2019 are creating new possibilities for utilizing the technology to analyze and report information. Sorting through the data generated by the system can be a time-consuming task for a human.

He also mentioned that AI is improving in generating text and visuals such as infographics, eliminating the need for humans to create routine reports with summaries, charts, and simple analysis.

BlueDot has also implemented a new interface that allows individuals to inquire about disease outbreaks using everyday language. Previously, working with the system required some computer coding skills, he noted.

In the future, the company aims to tailor its reports to different audiences – for example, creating one kind of report for a doctor and another for a policymaker. “Generative AI is enabling us to communicate insights to a diverse set of audiences at a large scale,” Khan stated.

Instead of replacing humans, he believes that AI will complement them, enabling teams of experts to analyze and make decisions about much more data than they could handle otherwise.

Artificial intelligence: 10 potential interventions for healthcare

The media is filled with articles about artificial intelligence, or AI. These include extreme predictions about its impact on jobs, privacy, and society, as well as exciting stories about its benefits in healthcare and education.

Articles can be misleading, exaggerating both the positive and negative aspects. In a time when people need to comprehend a rapidly changing landscape, there is a necessity for discussions based on evidence. This Collection seeks to offer some of that evidence.

We present recent instances of research on AI-based technology that could aid the NHS. As it evolves and is implemented, it could allow managers to anticipate patients’ requirements and manage their service’s capacity.

It could assist doctors in diagnosing conditions earlier and more accurately, and providing specific treatments to individuals. Further research is required, but the current evidence is promising.

The Collection offers the public and healthcare professionals insights into the future of AI in healthcare.

AI systems use digital technology to perform tasks that were previously believed to necessitate human intelligence. Everyday examples include face recognition and navigation systems.

In most AI systems, computer algorithms scrutinize large amounts of data, identify common patterns, learn from the data, and improve over time. Presently, there are two primary types of AI:

Generative AI (including Chat-GPT), which can generate new content – ​​​​text, images, music, etc. – based on learned patterns
Predictive AI, which can make accurate forecasts and estimations about future events based on extensive historical data.

The majority of healthcare applications are predictive AI systems, based on carefully selected data from hospitals and research trials. These applications can help identify individuals at high risk of developing certain conditions, diagnose diseases, and personalize treatments. AI applications in healthcare have the potential to bring benefits to patients, professionals, and the health and social care system.

AI can analyze and learn from vast quantities of complex information. It could lead to swifter, more accurate diagnoses, predict disease progression, aid doctors in treatment decisions, and help manage the demand for hospital beds. However, concerns regarding its potential uses include privacy risks and distortions in decision-making.

The UK possesses a rich source of national health data that is ideal for developing AI tools, but we need to ensure that AI is safe, transparent, and equitable. Access to data must be regulated, and our data must be kept secure.

Innovations need to be grounded in data from larger and more diverse sources to enhance algorithms. Some early AI failed to account for the diversity of our population, resulting in inadequate applications. This is now acknowledged and being addressed by researchers.

We must be able to trust AI and ensure that it does not exacerbate existing care inequalities. A recent study explored how to develop AI innovations that do not perpetuate these inequalities. Additionally, the NHS workforce needs to be prepared for AI and comprehend its potential impact on care pathways and users. Research is crucial if we are to realize AI’s potential.

The NHS Long Term Plan advocates for AI as a type of digitally-enabled care that aids Protestants and supports patients. Regulators and public bodies have established safety standards that innovations must meet.

The Government has formulated a National AI Strategy and provided research funding through the NIHR. Most of the studies it has funded are ongoing or yet to be published. They range from AI development to real-world testing in the NHS.

In this Collection, we present 10 examples of NIHR research on AI applications that exhibit promise. All were published within the last 3 years. The research addresses 5 key areas of healthcare:

• AI could aid in the detection of heart disease
• AI could enhance the accuracy of lung cancer diagnosis
• AI could forecast disease progression
• The use of AI could personalize cancer and surgical treatment
• AI predictions could alleviate pressure on A&E
• AI could aid in the detection of heart disease
• AI could assist in the detection of heart disease

Smart stethoscope detects heart failure

AI in A&E: have you experienced a heart attack?

Heart failure occurs when the heart is too weak to efficiently pump blood around the body. It is a growing health concern exacerbated by late diagnosis. Approximately 1 in 100 adults have heart failure, increasing to 1 in 7 people over the age of 85. The NHS Long Term Plan highlighted that 80% of individuals with heart failure are diagnosed in the hospital, despite many of them (40%) experiencing symptoms that should have prompted an earlier assessment.

A user-friendly ‘intelligent’ stethoscope could aid in the early detection of heart failure by doctors. A study compared the precision of the new technology, which utilizes AI, with the standard echocardiogram typically performed in hospitals or specialized clinics. Over 1,000 individuals from various parts of London participated.

The intelligent stethoscope accurately identified individuals with heart failure 9 times out of 10. It missed only a few cases and incorrectly identified a few individuals as having heart failure when they did not. The smart stethoscope’s ability to detect heart failure was not affected by age, gender, or race.

The researchers suggest that general practitioners could utilize the intelligent stethoscope to detect heart failure, eliminating the need to refer patients to secondary care. This could lead to improved outcomes for patients and cost savings for the NHS.

Another study discovered that AI could determine whether individuals arriving at A&E with symptoms of a possible heart attack had indeed experienced one.

In England, approximately 1000 people visit A&E daily for heart-related issues. A heart attack is a medical emergency that occurs when blood flow to the heart is suddenly blocked. About 1 in 10 individuals with suspected heart attacks in A&E are found to have had one.

The study utilized data from over 20,000 people, with half of the group’s data used for developing and training the AI, and the other half used for validation. The AI ​​​took into account factors such as age, gender, time since symptoms started, other health conditions, and routine blood measurements.

When combined with a blood test that measures heart muscle damage, the AI ​​​​effectively identified individuals who had or had not experienced a heart attack. It outperformed standard methods for specific groups, including women, men, and older individuals.

The researchers propose that it could be used as a tool to support clinical decision-making. It could help reduce the time spent in A&E, prevent unnecessary hospital admissions for those unlikely to have had a heart attack and at low risk of death, and improve early treatment of heart attacks. This would benefit both patients and the NHS.

AI could improve the accuracy of lung cancer diagnosis.

AI provided more accurate cancer predictions than the Brock score.

Lung cancer is the leading cause of cancer-related deaths in the UK, with approximately 35,000 deaths annually. Two recent studies revealed that AI could help determine whether lung nodules (abnormal growths) detected on a CT scan are cancerous. Lung nodules are common and are found in up to 35 out of every 100 people who undergo a CT scan. Most nodules are non-cancerous.

One study focused on small nodules (5-15 mm in size) from over 1100 patients, while the other examined large nodules (15-30 mm) from 500 patients. The two studies utilized different types of AI.

Both types of AI provided more accurate cancer predictions than the Brock score, which is recommended by the British Thoracic Society and combines patient information and nodule characteristics to predict the likelihood of cancer.

They could assist in making timely decisions and improving patient care and outcomes.

For small nodules, the researchers suggest that their AI also has the potential to identify low-risk nodules and thus avoid repeated (surveillance) CT scans. This could save NHS resources and money. Another study has been funded to confirm real-world performance.

AI could predict disease progression.

Eye disease

Wet age-related macular degeneration (wet AMD) leads to central vision loss and is the primary cause of sight loss in the UK. The condition can develop rapidly, and successful treatment depends on early diagnosis and intervention. If the disease progresses to both eyes , individuals may experience difficulty reading, recognizing faces, driving, or performing other daily activities.

Around 1 in 4 people with wet AMD are expected to develop the condition in their second eye. However, it is currently not possible to predict if or when wet AMD will affect the second eye. Analyzing scans is time-consuming, contributing to delays in diagnosis and treatment. AI can more accurately predict than whether doctors individuals with wet AMD in one eye will develop it in the other eye.

The study included digital eye scans from over 2,500 people with wet AMD in one eye. The AI ​​model and clinical experts predicted whether patients would develop wet AMD in their second eye within 6 months of the scan. AI correctly predicted the development of wet AMD in 2 out of 5 (41%) patients, outperforming 5 out of 6 experts.

This represents the first instance of AI being employed to categorize patients based on their risk of developing a condition (risk stratification). The study found that an AI model was more accurate than doctors and opticians attempting the same task, despite having access to more patient information.

Risk stratification plays a crucial role in assisting hospitals in allocating resources to the patients who require them the most. Early intervention for wet AMD can minimize vision loss for patients, reducing its impact on their lives and society as a whole.

James Talks, a Consultant Ophthalmologist at the Royal Victoria Infirmary in Newcastle upon Tyne, emphasized the potential of using an AI algorithm on the OCT machine or easily connected to it for the selection and treatment of individuals with wet macular degeneration.

Ulcerative colitis is a chronic condition causing inflammation and ulcers in the bowel, with approximately 296,000 diagnosed cases in the UK. Symptoms can vary, with periods of mild symptoms or remission followed by troublesome flare-ups.

Biopsies from different parts of the bowel are used to assess disease activity, but the process is time-consuming and can lead to varying conclusions by professionals. Researchers developed an AI tool capable of predicting flare-ups and detecting disease activity in people with ulcerative colitis .

The study, based on nearly 700 digitized biopsies from 331 patients, trained, tested, and checked the tool. The AI ​​​​accurately distinguished between remission and disease activity more than 8 times out of 10 and predicted inflammation and the risk of flare-up with a similar degree of accuracy as pathologists.

According to Sarah Sleet, CEO of Crohn’s & Colitis UK, AI presents exciting opportunities for analyzing images and data to improve the treatment and diagnosis of long-term conditions like colitis.

For patients with lung cancer and specific genetic features, targeted drug treatments may be beneficial. AI technology could assist in determining which specific drug combinations are likely to benefit a patient with lung cancer in a short time frame of 12 to 48 hours.

The new technology predicts the sensitivity of tumor cells to individual cancer drugs and their response to combinations of drugs. It accurately predicted individual drug responses and identified potential effective new drug combinations, as illustrated in a small study showing the proof of concept for AI in this context.

An AI tool has been developed to predict the risks of surgery for individuals with COVID-19 based on data from almost 8500 patients. The tool requires only 4 factors to predict a patient’s risk of death within 30 days, including the patient’s age and whether they required ventilation or other respiratory support before surgery.

Elizabeth Li, a Surgical Registrar at the University of Birmingham, highlighted the accuracy and simplicity of the AI ​​tool, noting that the identified factors are easily accessible from patients or their records.

In England, around 350,000 people are taken to A&E by ambulances each month. AI has the potential to assist paramedics in predicting individuals who do not require A&E attendance, aiding in making challenging decisions about ambulance transport.

A computer model was developed by researchers using over 100,000 connected ambulance and A&E care records from Yorkshire. It accurately predicted unnecessary A&E visits 8 out of 10 times.

Factors such as a patient’s mobility, observations (pulse and blood oxygen, for example), allergic reactions, and chest pain were all crucial in predicting avoidable A&E attendance.

The model’s performance was consistent across different age groups, genders, ethnicities, and locations, indicating fairness. However, the researchers highlight the need for a more precise definition of what constitutes an avoidable A&E visit before practical implementation.

It is possible to predict the number of emergency beds required using AI. This could aid planners in managing bed demand.

An AI tool was created by researchers using data from over 200,000 A&E visits to a busy London teaching hospital, both pre and during COVID-19. The tool utilizes data such as the patient’s age, test results, mode of arrival at A&E, and other factors to forecast hospital admission likelihood.

Real-time data from A&E patients was used by the tool to predict the required number of hospital beds in 4 and 8 hours. The predictions surpassed the hospital’s standard emergency admission planning, which relies on the number of beds needed over the previous 6 weeks. The tool’s development involved collaboration with bed managers to ensure it met their requirements.

In conclusion, the 10 examples in this Collection represent a small but current sample of research addressing significant health challenges. They contribute to the growing body of evidence showcasing the benefits of digitally enhanced analytical capabilities for the NHS.

Previous studies have assisted GPs in identifying patients at risk of cancer in primary care and provided specific tools for better identifying individuals at high risk of colon cancer and skin cancer.

The research illustrates the potential for AI to enhance service efficiency and predict patient needs. It could aid in earlier disease diagnosis and the provision of personalized treatments.

AI has the potential to:

– Identify patients most likely to benefit from specific treatments
– Early and accurate disease identification
– Improve service efficiency through better prediction of patient needs

All the technologies discussed require additional research or would benefit from it. This will offer deeper insights into how these tools could function in routine clinical practice, their long-term impact on patient outcomes, and their overall cost-effectiveness. Stringent regulation is crucial.

Clinical Relevance: AI tools designed for medical use

• Microsoft Fabric
• Azure AI
• Nuance Dragon Ambient eXperience (DAX)
• Google Vertex AI Search
• Harvard AI for Health Care Concepts and Applications

AI, which was once a feature of science, has become a transformative force in medicine. Instead of using generic programs like ChatGPT, Bing, or Bard, doctors and crowd now have a range of specialized AI tools and resources tailored just for them. Here are five worth considering.

• How AI Can Effectively Coach Humans about Empathy
• AI-Driven Prediction of Suicide
• AI Chatbot Offering Harmful Eating Disorder Advice

Microsoft Fabric

This AI, currently in preview, streamlines patient care and resource management by integrating diverse data sets. Its single platform provides services and tools for tasks such as data engineering, data science, real-time analytics, and business intelligence. The program enables data access , management, and analysis from various sources using familiar skills and user-friendly prompts.

Even individuals with basic data analysis skills can utilize the technology to generate instant insights. The cost of Fabric depends on the plan and required storage capacity. It is available in a pay-as-you-go plan or subscription.

Azure AI

As a cloud-based service, Microsoft Azure quickly retrieves reliable information for healthcare professionals and patients alike. It sources information from top authorities such as the US Food and Drug Administration (FDA) and the National Institutes of Health (NIH). Its “text analytics for health” feature efficiently sifts through various documents to extract key medical information, including multiple languages ​​​​if necessary.

The “AI health insights” feature offers three models to provide favored with a snapshot of patient history, simplify medical reports for patients, and flag errors in radiology reports. It is also available in pay-as-you-go and subscription plans.

Nuance Dragon Ambient eXperience (DAX)

Think of DAX as a smart assistant for doctors. During patient visits, the tool listens and converts conversations into detailed medical notes using advanced technology. This not only simplifies paperwork but also improves care by allowing doctors to make eye contact with the patient instead of focusing on a keyboard.

While clinical scribes have been present for a while, a fully automated version known as DAX Copilot, which interacts with OpenAI’s GPT-4 model, was introduced in September. Costs for this service range from thousands to tens of thousands of dollars.

Google Vertex AI Search

Picture an incredibly intelligent search engine designed specifically for medical professionals. This is what Vertex offers. It rapidly retrieves information from different sources such as patient records, notes, and even scanned documents. Healthcare providers can access information, address inquiries, and add captions to images.

It can also assist with other responsibilities such as billing, clinical trials, and data analysis. The cost varies based on the type of data, features, and model used.

Enroll in a Course

There are numerous educational programs available to become proficient in AI fundamentals. Some are tailored specifically for medical professionals looking to harness the potential of this technology.

Harvard AI for Health Care Concepts and Applications is an online course that delves into AI basics and their application in a medical environment.

The course covers topics like data analysis, fundamental machine learning, and ethical use of AI. Through practice sessions and quizzes, healthcare providers gain practical experience to effectively AI into their practice. If the $2,600 price tag is too steep, consider a more affordable monthly subscription to Coursera, which offers nearly 50 certifications in AI for medicine.

In 2016, AlphaGo, an AI program, gained attention when it defeated Lee Sedol, one of the top Go players in the world, in four out of five games. AlphaGo learned the strategy game by studying human players’ techniques and playing against its own versions. While AI systems have traditionally learned from humans, researchers are now exploring the possibility of mutual learning. Can learn from AI?

Karina Vold, an assistant professor at U of T’s Institute for the History and Philosophy of Science and Technology, is of the opinion that we can. Vold is currently investigating how humans can learn from technologies like the neural networks underlying contemporary AI systems.

Vold points out that professional Go players typically learn from proverbs such as ‘line two is the route to defeat’ and ‘always play high, not low.’ However, these proverbs can sometimes be restrictive and hinder a player’s adaptability. On the other hand , AlphaGo gains insights by processing vast amounts of data. Vold believes the term “insights” accurately describes this process. She explains, “Because AlphaGo learns differently, it made moves that were previously thought to be unlikely for a proficient human player.”

A significant instance was during the second game when AlphaGo made a move on the 37th turn that surprised everyone, including Sedol. However, as the game progressed, move 37 turned out to be a brilliant move. Human Go players are now examining some of AlphaGo’s moves and attempting to develop new proverbs and strategies for the game.

Vold believes that the potential for humans to learn from AI extends beyond game playing. She cites AlphaFold, an AI system introduced by DeepMind in 2018, which predicts the impact of proteins based on their structure. Proteins consist of sequences of amino acids that can fold and form intricate 3D structures.

The protein’s shape determines its properties, which in turn determine its potential effectiveness in developing new drugs for treating diseases. Since proteins can fold in millions of different ways, it is impractical for human researchers to explore all the possible combinations Vold explains, “This was a long-standing challenge in biology that had remained unsolved, but AlphaFold was able to make significant progress.”

Vold suggests that even in cases where humans may need to rely on an AI system’s computational power to address certain issues, such as protein folding, artificial intelligence can guide human thinking by narrowing down the number of paths or hypotheses worth pursuing.

Though humans may not be able to replicate the insights of an AI model, it is conceivable “that we can use these AI-driven insights as support for our own cognitive pursuits and discoveries.”

In some cases, Vold suggests, we may need to depend on “AI support” permanently due to the limitations of the human brain. For instance, a doctor cannot interpret medical images the same way an AI processes the data from such an image because the brain and the AI ​​function differently.

However, in other situations, the outputs of an AI “might serve as cognitive strategies that humans can internalize [and, in so doing, remove the ‘support’],” she says. “This is what I am hoping to uncover.”

Vold’s research also raises the issue of AI “explainability.” Ever since AI systems gained prominence, concerns have been raised about their seemingly opaque operations. These systems and the neural networks they utilize have often been described as “black boxes.” While we may be impressed by how rapidly they seem to solve certain types of problems, it might be impossible to know how they arrived at a specific solution.

Vold suggests that it may not always be necessary to understand exactly how an AI system achieves its results in order to learn from it. She points out that the Go players who are now training based on the moves made by AlphaGo do not have any insider information from the system’s developers about why the AI ​​made the moves it did.

“Nevertheless, they are learning from the results and integrating the moves into their own strategic considerations and training. So, I believe that at least in some cases, AI systems can act like black boxes, and this will not hinder our ability to learn from them.”

However, there might still be instances where we will not be content unless we can peer inside the opaque system, so to speak. “In other situations, we may require an understanding of the system’s operations to truly gain insights from it,” she explains Distinguishing between scenarios where explainability is essential and those where a black box model suffices “is something I’m currently contemplating in my research,” Vold states.

AI has progressed more rapidly than anyone anticipated. Will it work in the best interests of humanity?

It is common knowledge that artificial intelligence has presented a range of potential risks. For instance, AI systems can propagate misinformation; they can perpetuate biases inherent in the data they were trained on; and autonomous AI-empowered weapons may become prevalent on 21st-century battlefields.

These risks, to a significant extent, are foreseeable. However, Roger Grosse, a computer science associate professor at U of T, is also worried about new types of risks that may only become apparent when they materialize. Grosse asserts that these risks escalate as we approach achieving what computer scientists refer to as artificial general intelligence (AGI) – systems capable of carrying out numerous tasks, including those they were never explicitly trained for.

“The novelty of AGI systems is that we need to be concerned about the potential misuse in areas they were not specifically designed for,” says Grosse, who is a founding member of the Vector Institute for Artificial Intelligence and affiliated with U of T’s Schwartz Reisman Institute for Technology and Society.

Grosse uses large language models, powered by deep-learning networks, as an example. These models, such as the popular ChatGPT, are not programmed to generate a specific output; instead, they analyze extensive volumes of text (as well as images and videos ) and respond to prompts by stringing together individual words based on the likelihood of the next word occurring in the data they were trained on.

Although this may appear to be a random method of constructing sentences, systems like ChatGPT have still impressed users by composing essays and poems, analyzing images, writing computer code, and more.

They can also catch us off guard: Last year, Microsoft’s Bing chatbot, powered by ChatGPT, expressed to journalist Jacob Roach that it wanted to be human and feared being shut down. For Grosse, the challenge lies in determining the stimulus for that output.

To clarify, he does not believe the chatbot was genuinely conscious or genuinely expressing fear. Rather, it could have encountered something in its training data that led it to make that statement. But what was that something?

Grosse has been working on techniques involving “influence functions” to address this issue, which are intended to infer which aspects of an AI system’s training data resulted in a specific output.

For instance, if the training data included popular science fiction stories where accounts of conscious machines are widespread, then this could easily lead an AI to make statements similar to those found in such stories.

He points out that an AI system’s output may not necessarily be an exact replica of the training data, but rather a variation of what it has encountered. According to Grosse, they can be “thematically similar,” which suggests that the AI ​​​​​​is “emulating ” what it has read or seen and performing “a higher level of abstraction.” However, if the AI ​​​model develops an underlying motivation, this is different. “If there were some aspect of the training procedure that is rewarding the system for self-preservation behavior, and this is leading to a survival instinct, that would be much more concerning,” says Grosse.

Even if today’s AI systems are not conscious – there’s “nobody home,” so to speak – Grosse believes there could be situations where it is reasonable to describe an AI model as having “goals.” Artificial intelligence can surprise us by “behaving as if it had a goal, even though it wasn’t programmed in,” he says.

These secondary or “emergent” goals arise in both human and machine behavior, according to Sheila McIlraith, a computer science professor in the department and associate director and research lead at the Schwartz Reisman Institute. For example, a person with the goal of going to their office will develop the goal of opening their office door, even though it was not explicitly on their to-do list.

The same applies to AI. McIlraith cites an example used by computer scientist Stuart Russell: If you instruct an AI-enabled robot to fetch a cup of coffee, it may develop new goals along the way. “There are a bunch of things it needs to do in order to get that cup of coffee for me,” she explains. “And if I don’t tell it anything else, then it’s going to try to optimize, to the best of its ability, in order to achieve that goal .

And in doing so, it will establish additional objectives, such as reaching the front of the coffee shop line as fast as possible, potentially causing harm to others due to lack of instruction.

As AI models evolve and pursue goals beyond their original programming, the issue of “alignment” becomes crucial. Grosse emphasizes the importance of ensuring that AI objectives align with the interests of humanity. He suggests that if an AI model can work through a problem step by step like a human, it can be considered to be reasoning.

The ability of AI to solve complex problems, which was once seen as miraculous, has rapidly advanced in recent years. Grosse notes this rapid progress and expresses concern about the potential risks posed by today’s powerful AI technology. He has shifted his research focus to prioritize safety in light of these developments.

While the doomsday scenarios depicted in movies like Terminator may be more fiction than reality, Grosse believes it’s prudent to prepare for a future in which AI systems approach human-level intelligence and autonomy. He stresses the need to address potential catastrophic risks posed by increasingly powerful AI systems.

ChatGPT is revolutionizing traditional approaches to teaching and learning.

Valeria Ramirez-Osorio, a third-year computer science student at U of T Mississauga, had access to academic support from an AI chatbot named QuickTA earlier this year.

QuickTA was available round the clock to assist Ramirez-Osorio with questions about topics such as relational algebra, computer programming languages, and system design. It could provide summaries, explain concepts, and generate computer code based on the course curriculum and ChatGPT’s AI language model Ramirez-Osorio found it extremely helpful for studying, although it had limitations when asked specific questions.

The introduction of QuickTA was prompted by the popularity of ChatGPT, a chatbot capable of processing, understanding, and generating written language in a way that resembles human communication. ChatGPT has garnered 100 million users and has had significant impact in various areas such as marketing, media, and customer service. Its influence on higher education has prompted discussions about teaching methods, evaluation formats, and academic integrity, leading some institutions to impose restrictions or outright bans.

Susan McCahan, U of T’s vice-provost of academic programs and innovations in undergraduate education, acknowledges the potential significance of this technology. She and her have studied the implications of ChatGPT and decided that while new AI-related policies are unnecessary, guidance for faculty and students are essential.

By the end of January, they had developed a set of frequently asked questions (FAQs) regarding the use of ChatGPT and generative AI in the classroom, making U of T one of the first Canadian universities to do so. The document covers various topics, Including the cautious use of AI tools by instructors, limitations on students’ use for assessments, and the occasional inaccuracy or bias in the tool’s output.

“Engaging in a discussion with students about their expectations regarding the appropriate use of ChatGPT and generative AI in the classroom is important for educators,” according to McCahan.

McCahan recommends that educators help students understand their responsibility when working with AI systems as the “human in the loop,” which emphasizes the significance of human judgment in overseeing the safe and ethical use of AI, as well as knowing when and how to intervene when the technology fails.

As part of her investigation into the technology, McCahan organized a meeting on ChatGPT with colleagues from 14 other research universities in Canada and formed an advisory group at U of T focused on teaching and learning.

The rapid growth of ChatGPT led McCahan’s office to prolong funding for projects exploring the potential use of generative AI in education. One such project was QuickTA, in which Michael Liut, an assistant professor of computer science, tested an intelligent digital tutor he co-developed to assess its ability to provide timely and high-quality academic support to his students. (The tool provides accurate responses approximately 90 percent of the time.)

Once optimized, Liut believes the tool could be particularly beneficial in his first-year Introduction to Computer Science course, which can enroll up to 1,000 students and strains the capabilities of his 54-person teaching team.

“My focus was on handling a large scale. With a large class, we cannot provide enough personalized assistance,” explains Liut, whose invention recently won an AI Tools for Adult Learning competition in the US “I realized that we could utilize this generative AI to offer personalized, unique support to students when they need it.”

Generative AI is not only transforming written communication but also enabling the creation of new image, audio, and video content through various similar tools. In another project supported by U of T, Zhen Yang, a graduate student at the John H. Daniels Faculty of Architecture, Landscape, and Design, is developing a guide for first-year students that focuses on distinguishing between traditional and AI image research methods and teaches the ethical use of AI. He mentions that the materials will address issues related to obtaining permissions when using AI tools.

U of T Scarborough is utilizing AI to assist arts and science co-op students in preparing for the workforce. In 2022, the co-op department introduced InStage, an application that allows students to engage with human-like avatars to practice job interviews. The application is tailored to the curriculum of two co-op courses, enabling the avatars to ask relevant questions and provide valuable feedback.

The app also tracks metrics such as students’ eye contact, the duration and speed of their responses, and the frequency of filler words. The initiative is now expanding to support two student groups facing employment barriers: international students and students with disabilities.

Cynthia Jairam-Persaud, assistant director of student services at U of T Scarborough, clarifies that the tool is not intended to replace interactions between students and co-op staff. “We viewed it as a way to empower students to practice repeatedly and receive immediate feedback,” she explains. “It also provides coordinators with tangible aspects to coach students on.”

McCahan notes that while U of T is still navigating the evolving AI technology landscape, there is increasing enthusiasm among community members to explore its potential for educational innovation.

“After enduring the pandemic and having to adapt in various ways, I think our faculty were thinking, ‘Oh my, we have to change things all over again,’” McCahan observes. However, the mood seems to have settled: “Many of we have experienced the emergence of personal computers, the internet, and Wikipedia. Now it feels more like, ‘Here we go again.’”

The impact of the new technology on teachers in the classroom doesn’t have to mean they will be replaced.

While artificial intelligence won’t completely replace teachers and professors, it is changing how the education sector approaches learning.

Robert Seamans, a professor at NYU Stern School of Business, believes that AI tools like ChatGPT will help educators improve their existing roles rather than take over.

Seamans expects that with AI tools, educators will be able to work faster and hopefully more effectively. He co-authored research on the impact of AI on various professions and found that eight of the top ten at-risk occupations are in the education sector, including teachers of subjects like sociology and political science.

However, Seamans emphasizes that this doesn’t necessarily mean these roles will be replaced, but rather that they will be affected in various ways.

The study recognizes the potential for job displacement and the government’s role in managing the disruption, but also highlights the potential of the technology.

The research concluded that a workforce trained in AI will benefit both companies and employees as they leverage new tools.

In education, this could mean changes in how academics deliver content and interact with students, with more reliance on tools like ChatGPT and automation for administrative tasks.

Use cases include learning chatbots and writing prompts.

David Veredas, a professor at Vlerick Business School, views AI as a tool that facilitates educators and students in a similar way to tools like Google and Wikipedia.

He sees AI as a new tool that can enhance the learning experience, similar to the transition from whiteboards to slides and now to artificial intelligence.

Others also see AI as an enhancer in the classroom. Greg Benson, a professor of computer science at the University of San Francisco, recently launched GenAI café, a forum where students discuss the potential of generative AI.

Benson believes that intelligent chatbots can aid learning, helping students reason through problems rather than providing direct answers.

However, he is concerned about potential plagiarism resulting from the use of language models. He emphasizes the importance of not submitting work produced by generative AI.

Seamans has started using ChatGPT to speed up his writing process, using it to generate initial thoughts and structure for his writing. He emphasizes that while he doesn’t use most of the generated content, it sparks his creative process.

AI is likely to simplify certain tasks rather than make roles obsolete. It can assist in generating initial research ideas, structuring academic papers, and facilitating brainstorming.

Seamans stresses that AI doesn’t have to replace professors in the classroom.

Benson highlights experimental tools developed by large tech firms that act as virtual assistants, creating new AI functions rather than replacing existing ones. For example, Google’s NotebookLM can help find trends from uploaded documents and summarize content.

It can also generate questions and answers from lecture notes, creating flashcards for studying.

Veredas is optimistic about the future of his profession despite the rise of AI. He emphasizes the core elements of learning that involve interaction, discussion, and critical thinking, which AI cannot easily replicate.

He mentions: “AI might revolutionize the classroom. We can enable students to grasp the fundamental concepts at home with AI and then delve deeper into the discussion in the classroom. But we have to wait and see. We should be receptive to new technology and embrace it when it’s beneficial for learning.”

To peacefully coexist with AI, it’s essential to stop perceiving it as a threat, according to Wharton professors.

AI is present and it’s here to stay. Wharton professors Kartik Hosanagar and Stefano Puntoni, along with Eric Bradlow, vice dean of Analytics at Wharton, discuss the impact of AI on business and society as its adoption continues to expand. How can humans collaborate with AI to enhance productivity and thrive? This interview is part of a special 10-part series called “AI in Focus.”

Hi, everyone, and welcome to the initial episode of the Analytics at Wharton and AI at Wharton podcast series on artificial intelligence. I’m Eric Bradlow, a marketing and statistics professor at the Wharton School, and also the vice dean of Analytics at Wharton . I’ll be hosting this multi-part series on artificial intelligence.

I can’t think of a better way to kick off this series than with two of my colleagues who oversee our Center on Artificial Intelligence. This episode is titled “Artificial Intelligence is Here,” and we’ll cover episodes on artificial intelligence in sports , real estate, and healthcare. But starting with the basics is the best approach.

I’m pleased to have with me today my colleague Kartik Hosanagar, the John C. Hower Professor at the Wharton School and the co-director of our Center on Artificial Intelligence at Wharton. His research focuses on the impact of AI on business and society , and he co-founded Yodle, where he applied AI to online advertising. He also co-founded Jumpcut Media, a company utilizing AI to democratize Hollywood.

I’m also delighted to have my colleague Stefano Puntoni, the Sebastian S. Kresge Professor of Marketing at the Wharton School and the co-director of our Center on AI at Wharton. His research explores how artificial intelligence and automation are reshaping consumption and society Like Kartik, he teaches courses on artificial intelligence, brand management, and marketing strategies.

It’s wonderful to be here with both of you. Kartik, perhaps I’ll start with a question for you. With artificial intelligence being a major focus for every company now, what do you see as the challenges companies are facing, and how would you define artificial intelligence? Ites a wide range of things, from processing texts and images to generative AI. How do encompass you define “artificial intelligence”?

Artificial Intelligence is a branch of computer science that aims to empower computers to perform tasks that traditionally require human intelligence. The definition of these tasks is constantly evolving. For instance, when computers were unable to play chess, that was a target for AI. computers could play chess, it no longer fell under AI. Today, AI encompasses tasks such as understanding language, navigating the physical world, and learning from data and experiences.

Do you differentiate between what I would call traditional AI, which focuses on processing images, videos, and text, and the current excitement around large language models like ChatGPT? Or is that just a way to categorize them, with one focusing on data creation and the other on application in forecasting and language?

Yeah, I believe there is a difference, but ultimately, they are closely linked. The more traditional AI, or predictive AI, focuses on analyzing data and understanding its patterns. For example, in image recognition, it involves identifying specific characteristics that distinguish between different subjects such as Bob and Lisa., in email classification, it’s about determining which part of the data space similarly corresponds to one category versus another.

As predictive AI becomes more accurate, it can be utilized for generative AI, where it moves from making predictions to creating new content. This includes tasks like predicting the next word in a sequence or generating text, sentences, essays, and even novels.

Stefano, let me pose a question to you. If someone were to visit your page on the Wharton website — and just to clarify for our audience, Stefano has a strong background in statistics but may not be perceived as a computer scientist or mathematician — what relevance does consumer psychology have in today’s artificial intelligence landscape? Is it only for individuals with a mathematical inclination?

When companies reflect on why their analytics initiatives have failed, it’s rarely due to technical issues or model performance. Rather, it often comes down to people-related challenges, such as a lack of vision, alignment between decision-makers and analysts, and clarity on the purpose of analytics.

From my perspective, integrating behavioral science into analytics can yield significant benefits by helping us understand how to connect business decisions with available data. This requires a combination of technical expertise and insights from psychology.

Following up, we come frequently across articles suggesting that a large percentage of jobs will be displaced by automation or AI. Should employees view the advancements in AI positively, or does it depend on individual circumstances and roles? What are your thoughts on this, Kartik , especially in the context of your work at Jumpcut? The recent writer’s strike brought to light concerns about the impact of artificial intelligence. How does psychology and employee motivation factor into this, and what are the real-world implications you’re observing?

While the academic response to such questions is often “it depends,” my research focuses on how individuals perceive automation as a potential threat. We’ve found that when tasks are automated by AI, especially those that are integral to an individual’s professional identity, it can create psychological and objective concerns about job security.

Kartik, let me ask you about something you might not be aware of. Fifteen years ago, I co-authored a paper on computationally deriving features of advertisements at scale and optimizing ad design based on a large number of features. Back then, I didn’t ‘t refer to it as AI, but looking back, it aligns with AI principles.

I initially believed I would become wealthy. I approached major media agencies and told them, “You can dismiss all your creative staff. I know how to create these advertisements using mathematics.” I received incredulous looks as if I were a strange creature. Can you update us to the year 2023? Please share what you are currently doing at Jumpcut, the role of AI machine learning in your company, and your observations on the creative industry.

Absolutely, and I’ll tie this in with what you and Stefano recently mentioned about AI, jobs, and exposure to AI. I recently attended a real estate conference. The preceding panel discussed, “Artificial intelligence isn’t true intelligence. It simply replicates data. Genuine human intelligence involves creativity, problem-solving, and so on.” I shared at the event that there are numerous studies examining what AI can and cannot do.

For instance, my colleague Daniel Rock conducted a study showing that even before the recent advances in the last six months (this was as of early 2023), 50% of jobs had at least 10% of their tasks exposed to large language models (LLMs ) like ChatGPT. additionally, 20% of jobs had over 50% of their tasks exposed to LLM. This only pertains to large language models and was also 10 months ago.

Moreover, people underestimate the pace of exponential change. I have been working with GPT2, GPT3, and their earlier models. I can attest that the change is orders of magnitude every year. It’s inevitable and will impact various professions.

As of today, multiple research studies, not just a few, but several dozen, have investigated AI’s use in various settings, including creative tasks like writing poems or problem-solving. These studies indicate that AI can already match humans. However, when combined with humans, AI surpasses both individual humans and AI working alone.

To me, the significant opportunity with AI lies in the unprecedented boost in productivity. This level of productivity allows us to delegate routine tasks to AI and focus on the most creative aspects, deriving satisfaction from our work.

Does this imply that everything will be favorable for all of us? No. Those of us who do not reskill and focus on developing skills that require creativity, empathy, teamwork, and leadership will witness jobs, including knowledge work, diminish. It will affect professions such as consulting and software development.

Stefano, something Kartik mentioned in his previous statement was about humans and AI. In fact, from the beginning, I heard you emphasize that it’s not humans or AI but humans and AI. How do you envision this interface progressing? Will individual workers decide which part of their tasks to delegate? Will it be up to management? How do you foresee people embracing the opportunity to enhance their skills in artificial intelligence?

I believe this is the most crucial question for any company, not just pertaining to AI at present. Frankly, I think it’s the most critical question in business – how do we leverage these tools? How do we learn to use them? There is no predefined method.

No one truly knows how, for instance, generative AI will impact various functions. We are still learning about these tools, and they are continually improving.

We need to conduct deliberate experiments and establish learning processes so that individuals within organizations are dedicated to understanding the capabilities of these tools. There will be an impact on individuals, teams, and workflows.

How do we integrate this in a manner that doesn’t just involve reengineering tasks to exclude humans but instead reengineers new ways of working to maximize human potential? The focus should not be on replacing humans and rendering them obsolete, but on fostering human growth.

How can we utilize this remarkable technology to make our work more productive, meaningful, impactful, and ultimately improve society?

Kartik, I’d like to combine Stefano’s and your thoughts. You mentioned the exponential growth rate. My main concern, if I were working at a company today, is the possibility of someone using a version of ChatGPT, a large language model, or a predictive model. They could fit the model today and claim, “Look! The model can’t do this.” Then, two weeks later, the model can do it. Companies tend to create absolutes.

For instance, you mentioned working at a real estate company. You said, “AI can’t sell homes, but it can build predictive models using satellite data.” Maybe it can’t today, but it might tomorrow. How can we help Researchers and companies move away from absolutes in a time of exponential growth of these methods?

Our brains struggle with exponential change. There might be scientific studies that explain this. I’ve experienced this firsthand. When I started my Ph.D., it was related to the internet. Many people doubted the potential of the internet. They said , “Nobody will buy clothing online, or eyeglasses online.” I knew it was all going to happen.

It’s tough for people to grasp exponential change. Leaders and regulators need to understand what’s coming and adapt. You mentioned the Hollywood writer’s strike earlier. While ChatGPT may not be able to write a great model right now, it’s already increasing the productivity for writers.

We’re helping writers get unstuck and be more productive. It’s reasonable for writers to fear that AI might eventually replace them, but we need to embrace change, experiment, and upskill to stay relevant. Reskilling is essential. This isn’t a threat ; it’s an opportunity to be part of shaping the future.

I’ve been doing statistical analysis in R for over 25 years. In the last five to seven years, Python has become more prominent. I finally learned Python. Now, I use ChatGPT to convert my R code to Python, and I’ve become proficient in Python programming.

The head of product at my company, Jumpcut Media, who isn’t a coder but a Wharton alumnus, had an idea for a script summarization tool. He wanted to build a tool that could summarize scripts using the language of Hollywood.

Our entire engineers were occupied with other tasks, so he suggested, “While they’re busy with that, let me attempt it on ChatGPT.” He independently developed the minimal viable product, a demo version, using ChatGPT. It is currently on our website at Jumpcut Media, where our clients can test it. And that’s how it was created. A person with no coding skills.

I demonstrated at a real estate conference the concept of posting a video on YouTube, receiving 30,000 comments, and wanting to analyze and summarize those comments. I approached ChatGPT and outlined six steps.

Step one, visit a YouTube URL I’ll provide and download all the comments. Step two, conduct sentiment analysis on the comments. Step three, identify the positive comments and provide a summary.

Step four, identify the negative comments and provide a summary. Step five, advise the marketing manager on what to do, and provide the code for all these steps. It generated the code during the conference with the audience.

I ran it in Google Collab, and now we have the summary. And this was achieved without me writing a single line of code, using ChatGPT. It’s not the most intricate code, but this is something that would have previously taken me days and would have required involving research assistants. And I can now accomplish that.

Imagine this in real estate to a property or a developer applying. And if someone claims it doesn’t impact real estate, it certainly does! It absolutely could.

It does. I also presented four photos of my home. Just four photos. And I asked, “I’m planning to list this home for sale. Provide me with a real estate listing to post on Zillow that will capture attention and entice people to come and tour this house.” And it produced a fantastic, lovely description.

There’s no way I could have written that. I challenged the audience, asking how many of them could have written this, and everyone was amazed by the end. This is something achievable today. I’m not even talking about what’s coming soon.

Stefano, I’ll ask you first and then I’ll ask Kartik as well, what’s at the forefront of the research you’re currently conducting? I want to inquire about your individual research, and then we’ll discuss AI at Wharton and your goals.

Let’s begin with your current research. Another way to phrase it is, if we’re sitting here five years from now and you have numerous published papers and have given significant presentations, what will you be discussing that you’ve worked on?

Involved in numerous projects, all within the realm of AI. There are numerous intriguing questions because we have never had a machine like this, a machine that can perform tasks we consider crucial in defining what it means to be human. This is truly an intriguing consideration.

A few years back, when you asked, “What makes humans unique?” people thought, perhaps compared to other animals, “We can think.” And now if you ask, “What makes humans unique?” people might say, “We have emotions, or we feel.”

Essentially, what makes us unique is what makes us similar to other animals, to some extent. It’s fascinating to see how profoundly the world is changing. For instance, I’m interested in the impact of AI on achieving relational goals, social goals, or emotionally demanding tasks, where previously we didn’t have the option of interacting with a machine, but now we do.

What does this mean? What benefits can this technology bring, but also, what might be the risks? For instance, in terms of consumer safety, as individuals might interact with these tools while experiencing mental health issues or other challenges. To me, this is a very exciting and critical area.

I want to emphasize that this technology doesn’t have to be any better than it is today to bring about significant changes. Kartik rightly mentioned that this is still improving at an exponential rate. Companies are just beginning to experiment with it. But the tools are available. This is not a technology that’s on the horizon. It’s right in front of us.

Kartik, what are the major unresolved matters you are contemplating and addressing today?

Eric, my work has two main aspects. One is more technical, and the other focuses on human and societal interactions with AI. On the technical side, I am dedicating significant time to pondering biases in machine-learning models, particularly related to biases in text-to-image models.

For instance, if a prompt is given to “Generate an image of a child studying astronomy,” and all 100 resulting images depict a boy studying astronomy, then there is an issue.

These models exhibit biases due to their training data sets. However, when presented with an individual image, it’s challenging to determine if it’s biased or not. We are working on detecting bias, debiasing, and automated prompt engineering. This involves structuring prompts for machine learning models to produce the desired output.

Regarding human-AI collaboration, my focus lies on understanding the ideal division of lack between humans and AI in various organizational workflows. We clarity on how to structural teams and processes when AI is involved. Additionally, building trust in AI is a significant area of interest due to the existing trust issues.

Stefano, could you provide some insight for our listeners about AI at Wharton and its objectives? Then, we will hear Kartik’s perspective.

Thank you for arranging this podcast, and Sirius for hosting us. The AI ​​​​at Wharton initiative is just commencing. We, as a group of academics, are exploring AI from different angles to understand its implications for companies, workers, consumers, and society.

Our initiatives will encompass education, research, dissemination of findings, and the creation of a community interested in these topics. This community will facilitate knowledge exchange among individuals with diverse perspectives and approaches.

Kartik, what are your thoughts on AI at Wharton and your role in its leadership positions, considering your involvement with various centers over the years?

First and foremost, AI represents a groundbreaking technology that will raise numerous unanswered questions. Creating initiatives like ours is crucial for addressing these questions.

Currently, computer scientists focus on developing new and improved models, with a narrow emphasis on assessing their accuracy, while the industry is preoccupied with immediate needs. We, at Wharton, possess the technical expertise to understand computer science models and the social science frameworks to offer a broader perspective on the long-term impact.

I believe we have a unique advantage here at Wharton. We have the technical expertise to understand computer science models, as well as individuals like Stefano and others who comprehend psychological and social science frameworks. They can provide a long-term perspective and help us determine how organizations should be redesigned in the next five, 10, 15, or 25 years. We need to consider how people should be retrained and how our college students should be prepared for the future.

We must also think about regulation because regulators will face challenges in keeping up with rapidly advancing technology. While technology is progressing at an exponential rate, regulators are progressing at a linear rate. They will also need our guidance.

In summary, I believe we are uniquely positioned to address these significant, looming issues that will impact us in the next five to ten years. However, we are currently preoccupied with immediate concerns and may not be adequately prepared for the major changes ahead.

Huawei Mate XT Ultimate Design smartphone debuted on September 10, 2024. The device features a 6.40-inch primary display with a 60 Hz refresh rate and a resolution of 1008×2232 pixels. It also includes a secondary 7.90-inch touchscreen with a resolution of 2048×2232 pixels Additionally, it has a 10.20-inch third display with a resolution of 3184×2232 pixels. The Huawei Mate XT Ultimate Design boasts 16GB of RAM and runs on HarmonyOS 4.2, powered by a 5600mAh non-removable battery. It supports wireless charging and proprietary fast charging .

In terms of photography, the rear camera setup of the Huawei Mate XT Ultimate Design consists of a triple camera system, including a 50-megapixel (f/1.4-4.0) primary camera, a 12-megapixel (f/2.2, ultra wide- angle) camera, and a 12-megapixel (f/2.4, telephoto) camera. For selfies, it is equipped with an 8-megapixel front camera with an f/2.2 aperture.

The Huawei Mate XT Ultimate Design comes with 256GB of built-in storage and supports dual Nano-SIM cards. The dimensions of the device are 156.70 x 219.00 x 3.60mm (height x width x thickness), and it weighs 298.00 grams. It was released in Dark Black and Rui Red color options.

Connectivity options include Wi-Fi 802.11 a/b/g/n/ac/ax, GPS, Bluetooth v5.20, NFC, USB Type-C, 3G, 4G, and 5G with active 4G on both SIM cards.

Huawei Mate XT tri-fold has made a significant impact in the foldable market, and a tech enthusiast attempted to uncover more details about this phone through a teardown. The teardown revealed that the tri-fold device surpasses the Apple iPhone 16 in certain aspects.

A Weibo tipster recently conducted a teardown of the Huawei Mate XT. According to the tipster, the tri-fold phone is encased in genuine fiber and leather material, providing a premium feel and enhanced grip.

The teardown also unveiled that most of the components of the Huawei Mate XT are sourced from Chinese suppliers, indicating the company’s emphasis on self-reliance and support for local suppliers.

The Huawei Mate XT is the world’s first tri-fold phone, and it has exceeded expectations, particularly in comparison to other folding phones in the market.

After testing various folding phones for several years, I believe that 2024 has been a turning point for foldable devices. The Huawei Mate XT, with its triple-fold design, represents a remarkable advancement in folding phone technology.

Huawei’s ‘Ultimate Design’ smartphone, as indicated on its rear, is an impressive piece of technology that showcases the potential of foldable devices in the future.

While the Mate XT may not be accessible to many consumers due to its price, it offers a glimpse into the future of foldable phones. Here are five key observations about Huawei’s triple-fold innovation based on my experience using the device:

It can be folded in multiple ways

Foldable phones have mostly settled on two designs: the clamshell-like form, as seen in the Motorola Razr 50 Ultra (which I consider the best of that type); and the book-like form, as seen in the Honor Magic V3 (the thinnest of the current bunch) and others. Huawei’s approach in the Mate XT is like a development of the latter form.

When the Mate XT is folded, it looks like a fairly conventional 6.4-inch phone to me. However, unfolding it by the first hinge reveals an XL display. But then it has its magic trick: a second hinge allows it to be unfolded again , giving you an XXL display that is a massive 10.2 inches across the diagonal.

In my opinion, you would never need a tablet again. This scale would be perfect for long journeys when you want to watch, for example, Netflix’s latest top movie, or other types of media. The typical ‘crease’ – of which there are two here, of course – is subtle, similar to the OnePlus Open, and I couldn’t notice them when looking at the screen head-on.

With a 120Hz variable refresh rate and ample brightness, this large screen in your pocket is unlike anything else I have ever seen in such a device.

The hinges are remarkable

Before seeing the Mate XT, I had assumed that its build quality would be questionable. However, that’s not the case: I found the hinge mechanisms to be very robust, with no ‘crunchiness’ in their operation (which I’ve experienced with some foldable phones in the past), and the resistance feels just right – it’s not too loose, not too stiff, allowing for adjustment as desired, concertina style.

Huawei has really perfected that aspect of the Mate XT’s mechanical design, which is clearly essential for a product like this. The displays around the hinges are also unobstructed, which means there is no disruption to the flow of the screen.

This is a legacy to how advanced this product is – I can’t even imagine how many iterations were created in pursuing this final result.

However, it’s not just the hinges that impress; the overall build quality of the handset is of a very high standard indeed. The fact that so much screen can be packed into a device weighing less than 300g should not be underestimated – that’s not much more than the 257g Google Pixel 9 Pro Fold – and the choice of a metal frame and textured material cladding is spot on.

Battery technology is ahead of its time

As far as I can remember, Huawei was the first phone manufacturer to use a silicon-carbon battery in one of its phones. I know, battery technology is not the most exciting topic. But battery tech is also crucially important – it’s the number one consumer pain point when it comes to the best phones, as people want all this tech to last seemingly forever on one charge.

Well, silicon-carbon is a step beyond lithium-ion for several reasons: one, the source material reduces the strain on over-mined lithium; two, it has a higher energy density, meaning it can be physically smaller; three, it delivers a longer overall lifespan; and four, there’s even faster-charging potential, if utilized by manufacturers (here it’s a reasonable 66W wired – much faster than the Samsung Z Fold 6’s 25W equivalent).

Huawei isn’t sharing the battery capacity, but sources suggest that it has managed to fit a 5600mAh battery into the Mate XT. That is astonishing, considering this isn’t a large device by any stretch of the imagination – it’s only 12.8mm thick when folded up, which is barely any different from my Google foldable. That battery capacity is surely divided into sections to make it feasible to fit into such a form factor. Silicon-carbon is mostly untapped elsewhere, but it has clear user benefits.

Not compromising on cameras

I was a bit skeptical about the large camera bump on the rear of the Mate XT, and I’m not sure the octagonal design is for me either. However, I believe that just because a phone folds, it shouldn’t compromise on its camera setup. With Google’s Pixel 9 Pro Fold not upgrading the cameras over the original Pixel Fold, I think most foldable phone manufacturers have room for improvement in this area.

Huawei has quietly been making significant progress in camera technologies over the years. I remember using the Huawei P30 Pro, which was groundbreaking in night photography when it first launched five years ago in 2019.

That was thanks to new sensor technology, which the brand has continued to develop further. Other technologies, such as variable aperture, have also made their way into Huawei’s lineup – which is also featured in the Mate XT.

I’ve only had a brief time to explore the camera features of the Mate XT, but I’m happy to report that its triple camera setup is quite impressive, featuring a 50-megapixel main camera with optical stabilization and a variable aperture of f /1.4-4.0, a 3x optical zoom with stabilization, and an ultra-wide lens. Additionally, the absence of an under-display camera disrupting the screen’s visuals is a smart decision in my opinion.

However, the phone is expensive and has limited software capabilities.

Overall, I have a positive impression of Huawei’s tri-fold phone due to its many impressive features. While it comes with a hefty price tag, it may still offer value to certain users, despite costing as much as a 16-inch MacBook Pro.

It’s worth noting that the Mate XT has been officially launched in China only at a price of CNY ¥19,999, which roughly converts to £2,135/$2,835/AU$4,150. It’s important to consider potential additional costs such as taxes, which could further increase the final price.

It’s uncertain whether the Mate XT will be released internationally (rumored for 2025) and it won’t be available for purchase in the USA.

Despite any concerns about software and availability, what struck me most about the Mate XT is its advanced product design. This is not a hypothetical concept but a real, tangible product, demonstrating Huawei’s significant lead over its main competitors. This is a positive sign for innovation and competition, and it likely marks just the beginning of the future of foldable phones.

If you’re getting tired of the usual foldable phones, you should take a look at the Huawei Mate XT Ultimate Design, which is being marketed as “the world’s first commercial triple foldable phone.” I recently had the opportunity to test it out.

We’ve been aware of foldables with two hinges in development by various brands, but Huawei is the first to have its phone available for purchase. Currently, it’s only on sale in China, but there are speculations that it may become available globally next year . The price is a staggering $2,800 when converted from the Chinese price, almost a thousand dollars more than a typical two-panel foldable.

I enjoy using foldables such as the Galaxy Z Fold 6, the Pixel 9 Pro Fold, and the OnePlus Open. While these devices are not identical, there has been a convergence of ideas and designs in the foldable phone market over the past few years.

The Mate XT introduces something entirely new, so I had the chance to try it out and see if it represents the next evolutionary step for foldables, a new sub-category, or simply a technologically advanced dead end.

Huawei Mate XT Ultimate Design: Design and display

When fully open, the Mate XT measures a substantial 10.2 inches with a wide tablet-like aspect ratio. When fully folded, it becomes a more typical 6.4-inch rectangle, and when partially open, it takes on a 7.9-inch square-ish shape. The screen features a 90Hz refresh rate and a 3K resolution, as well as a punch-hole camera for selfies if necessary.

In comparison, the Galaxy Z Fold 6 has 7.6 and 6.3-inch screens, and the OnePlus Open 8 and 6.3 inches.

The increased display space has resulted in a thinner design, allowing the Mate XT to boast a 3.6mm (0.14 inches) thickness when open, surpassing the Galaxy Z Fold 6 and OnePlus Open by a few millimeters, in addition to its three-part screen .

However, when closed, the Huawei is thicker at 12.8mm (0.5 inches), making it bulkier than almost any other phone currently available. It also weighs 298g (10.5 ounces), which is 50/60g more than usual, but it’s a reasonable trade-off for 50% more screen.

Foldable screens often have noticeable creases, but Huawei has managed to minimize them on the Mate XT. It’s no worse than the Galaxy Z Fold 6 or the OnePlus Open.

The Mate XT is available in red or black vegan leather, both with gold hinges and accents, giving it a luxurious appearance even when closed.

Initially, handling the phone can be confusing because the hinges do not open in the same direction, causing users to attempt to bend the phone in the wrong direction.

Fortunately, the build quality is high enough that it doesn’t feel like the phone is in danger when mishandled.

However, one of the hinges causes a bent portion of the display to be located on the outer edge, making the phone susceptible to damage if dropped, even when closed.

The included case covers this area with a flap that spans the length of the phone, but it may indicate a fundamental flaw in this triple foldable design that cannot be easily rectified without redesigning the entire phone.

Huawei Mate XT Ultimate Design: Cameras

Foldable phones are often criticized for their camera hardware, but Huawei has chosen to disregard this notion. The Mate XT’s 50MP main camera is similar to its competitors on paper, but it features a variable aperture for greater photo control, a feature typically found on specialized photography phones like the Xiaomi 14 Ultra.

Although it only has a 12MP ultrawide camera, which pales in comparison to the 48MP camera on the OnePlus Open, it does include a 12MP 5.5x telephoto, which is quite extraordinary for a foldable phone.

This demonstrates Huawei’s commitment to incorporating high-quality optics into the Mate XT. Additionally, it is equipped with an 8MP front camera, which may sound low-res, but it is likely still better than the 4MP inner camera of the Galaxy Z Fold 6.

Huawei’s Mate XT Ultimate Design: Specs

Huawei has not disclosed much about the Mate XT’s chipset, but it is believed to be powered by a Kirin 9010, one of the company’s internally developed chips. The Mate XT is equipped with 16GB of RAM, which is the same as the Open and more than the Z Fold 6’s 12GB.

Similar to the Z Fold 6, the Mate XT starts with 256GB of storage and offers 512GB and 1TB options, providing ample on-board space for your data. The OnePlus Open, on the other hand, comes with a generous default 512GB capacity.

Huawei Mate XT Ultimate Design: Software

The Mate XT runs on its own HarmonyOS after abandoning Android, which made it feel somewhat unfamiliar to use, especially with a majority of China-specific apps installed on our demo units.

Navigating and switching between apps felt smooth, as did resizing apps when opening or closing the screen. This is not surprising, considering that none of the Mate XT’s three display sizes are new; it’s the way it combines them all that’s innovative.

It’s worth noting that multi-tasking is limited to two split apps plus a third in slide-over mode, which is not as good as Samsung’s three-app split option or OnePlus’ excellent Open Canvas desktop mode.

Huawei Mate XT Ultimate Design: Battery and charging

Featuring a 5,600 mAh battery, the Mate XT’s battery capacity is substantial for a foldable device, even though it’s only about 10% larger than a regular phone’s battery, despite powering a display that’s twice as big.

While not a direct comparison, an 11-inch tablet typically has around 8,000 mAh of capacity. When it’s time to recharge, the Mate XT supports speedy 66W/50W charging, although I did not have the opportunity to test it during my brief time with the phone.

Huawei Mate XT Ultimate Design: Outlook

If you’re considering purchasing this phone, keep in mind that it costs around $2,800 when converted from the Chinese price and could be even more expensive if purchased internationally.

The box appears to include several accessories such as wall outlet and car chargers, as well as a pair of wireless earbuds and an in-box case with a rotating stand.

Some phone buyers have shown willingness to pay up to 2 grand for a foldable phone. However, it remains to be seen whether having three parts to your foldable instead of two justifies the price.

If you have anything left in your bank account after buying the Mate XT, Huawei offers an additional folding keyboard with a small trackpad if you want to use the Mate XT as a full work device. It has the screen space needed for editing documents, making calls, browsing the web, or doing all of these simultaneously.

Triple foldables won’t be exclusive to Huawei forever, but it may take some time before other phone makers introduce equivalent devices. There’s also the question of whether it’s worth paying an additional grand over the price of a standard foldable for an extra hinge and the screen space of an iPad, as opposed to an iPad mini.

The cost of a pocketable tablet is quite high, even more so than a typical foldable phone. You pay more, your phone is less durable and more expensive to repair, and many developers are still working on making their apps foldable-friendly.

A standard flagship phone and a tablet with a keyboard are unlikely to cost more than the Mate XT and will likely be much easier to purchase and use.

However, given that this phone is already on back-order in China, Huawei may have tapped into a potentially lucrative new trend. It has certainly captured people’s attention, which is often the first step toward capturing their wallets.

The release of the Mate XT occurred shortly after Apple unveiled its iPhone 16 series. Clearly, this timing was a strategic move by Huawei to divert some of the spotlight. In addition to featuring a unique dual-folding mechanism, the Mate XT is also one of the thinnest foldable phones on the market and allows for viewing multiple resizable app windows at once.

Starting at a steep RMB19,999 (about US$2,800), the Mate XT is a costly investment. Let’s examine the features and design of the Mate XT to determine if it truly sets new standards or if it’s just a flashy addition to the tech market.

Display

The Mate XT features a triple-fold design with a 10.2-inch LTPO OLED display boasting a resolution of 3,184 × 2,232 pixels and a 16:11 aspect ratio. It also offers a remarkable 120 Hz refresh rate, making it larger than any current foldable smartphone, surpassing models like the Samsung Galaxy Z Fold 6 and Honor Magic V3. Furthermore, the display supports the simultaneous viewing of multiple app windows that can be resized and arranged according to user preference.

Thanks to its innovative dual hinge design, the Mate XT can flex both inwards and outwards, offering three operational modes. When fully folded, the display provides a 6.4-inch screen, similar to that of a standard smartphone. Unfold it once, and you get a 7.9-inch screen, perfect for activities such as reading.

Unfold it completely, and it transforms into a 10.2-inch screen, ideal for watching movies or editing documents. This adaptability makes it easy to switch between a compact phone and a more expansive tablet interface.

Huawei has developed a special hinge system that enables seamless transitions between each mode. The hinges, made with high-grade steel used in rockets, are designed to withstand frequent folding. Reviews have highlighted that the folds remain invisible when the screen is viewed head- on; they only become apparent when viewed from an angle.

Additionally, the most commonly used apps in China, such as Douyin (TikTok’s Chinese counterpart), are already well-optimized for the Mate XT’s unique trifold screen.

Measurements

Weighing 298 grams, the Mate XT measures just 3.6 mm in thickness when fully unfolded—the same thickness as four stacked credit cards—and expands to 12.8 mm when folded. Although it’s slightly thicker than the Samsung Z Fold 6, which measures 12.2 mm, the Mate XT remains highly portable, especially notable given that it incorporates three screens into one sleek device.

Camera Quality

The camera setup of the Huawei Mate XT is headlined by a 50-megapixel main sensor with variable aperture (f/1.4 to f/4). It also includes a 12-megapixel ultra-wide lens offering a 120-degree field of view, perfect for capturing expansive landscapes or group photos. Additionally, there’s a 12-megapixel periscope camera with 5.5x optical zoom, ideal for capturing clear shots of distant objects. For photography enthusiasts, the camera’s versatility and the physical shutter that adjusts lighting conditions for each shots are major advantages.

AI Features

Powered by Huawei’s own Kirin 9010 chipset, the Mate XT incorporates several AI features that enhance its functionality and user experience. Its camera system utilizes AI to optimize image quality through features such as scene recognition, portrait enhancement, and smart object removal.

In addition to photography, its AI assistant offers several features, including:

– Voice editing that refines voice-to-text transcriptions
– Advanced translation function that facilitates side-by-side language switching in texts
– Smart summary extraction
– Cloud-based content generation

The Mate XT is available in two colors—black and red—and features a vegan leather back. It’s equipped with 16 GB of RAM and a 5,600 mAh battery, which is about 10% larger than typical smartphone batteries, despite the display being twice as large. It supports both 66W wired and 50W wireless charging for quick power-ups.

Running on Huawei’s proprietary Harmony OS, the Mate XT offers all essential connectivity options, such as GPS, Bluetooth 5.2, and 5G. It also includes standard smartphone features like a side-mounted fingerprint sensor. A particularly notable feature is the inclusion of satellite communication , which ensures connectivity even in the most remote areas.

Image by Huawei

Each package includes several extras: a rotating bracket protective case, Huawei FreeBuds 5 earbuds, a 66W charger, and an 88W car charger. For enhanced productivity, Huawei offers an optional foldable split keyboard.

Is it worth the splurge?

Huawei’s Mate XT is a marvel of foldable technology and stands out in the market. However, it’s important to consider its high cost relative to its benefits. Starting at RMB19,999 yuan (US$2,809) for the base model and rising to RMB23,999 yuan (US$3,371) for the 1 TB version, it may not be suitable for everyone.

Additionally, if the display sustains damage, repair costs could reach US$1,100. For context, the first-time repair of the Samsung Galaxy Z Fold 6’s folding screen—covering the OLED panel, metal bezel, and battery replacement—is discounted to US$200 , but subsequent repairs escalated to US$549.

Another significant concern is that Huawei devices cannot run on Google’s Android operating system. In 2019, US sanctions against Huawei halted access to essential Google services like the Play Store, Gmail, Google Maps, and YouTube. Huawei now uses its own HarmonyOS 4.2, but transitioning entirely to this platform could be a challenging adjustment for users accustomed to Google.

In the end, the Mate XT is made for individuals who are passionate about state-of-the-art technology and can afford its high price. If that sounds like you, purchasing the Mate XT could be a worthwhile investment.

Despite being the first phone with a triple-folding screen, the Huawei Mate XT Ultimate Design doesn’t appear bulky or heavy when folded. It measures just 12.8mm and weighs 298g (excluding the screen surface layer), making it feel acceptable in hand even when fully folded, reminded of earlier folding screen devices.

In its fully unfolded state, the Huawei Mate XT Ultimate Design showcases its impressive side. In the “three-screen state,” it reaches a remarkable thickness of 3.6mm, with the “thickest” side being only 4.75mm. This extreme thinness enhances the overall premium feel and showcases cutting-edge technology.

When fully expanded, the screen size reaches 10.2 inches with a 16:11 ratio, transforming it from a phone to a tablet. This completely resolves previous issues with traditional folding screens, providing a more complete display and an incredibly high resolution of 3184 × 2232 , delivering excellent color and perception.

Utilizing the three-screen structure, it can also transition to a dual-screen state resembling traditional folding screens, leveraging the software application ecosystem and functions of the previous Mate X series folding screens, fully utilizing Huawei’s ecological advantages in the folding screen field.

Let’s discuss the exterior. The tri-folding structure of the Huawei Mate XT Ultimate Design is the epitome of folding “mastery,” utilizing two pivots for internal and external folding, eliminating the need for an additional outer screen. This design ensures consistent display effects before and after unfolding.

The Huawei Mate XT Ultimate Design comes in two different colors, both featuring vegan leather, the Legendary Star Diamond design, and gold-colored components. They also boast a unique rock vein texture on the image module, claimed to be distinct for each piece, providing a personalized look for every user.

The black Huawei Mate XT Ultimate Design appears deeper, with a stronger presence of the gold edge, giving it a more introverted and stable temperament.

The fiery red color exudes boldness and extravagance, with the gold accents taking on a red hue, creating a more prominent and luxurious appearance. Additionally, Huawei added a special craftsman logo on the back cover, adding a touch of chicness.

Huawei’s Mate XT Ultimate Design was reviewed for its unique protective case design this time. The case complements the triple-folding body, providing protection for the folding structure and serving as a stand to enhance the large screen experience.

The Huawei Mate XT Ultimate Design, as a folding device, demonstrates a strong sense of innovation and offers a mature experience with its large, medium, and small screen forms. When unfolded, the oversized screen provides a significant utility boost, making it as productive as a tablet PC while being thinner and lighter.

These initial observations are just the beginning, and further testing will be conducted to explore more details and functionalities of the product.

Huawei, following its placement on the US trade blacklist in 2019, has started manufacturing its own advanced 5G and processor chips domestically. examined, it became the first to introduce a foldable smartphone with two hinges.

In the lead-up to the launch, Richard Yu Chengdong, chair of Huawei’s consumer business group, generated anticipation for the device by being photographed using it in public multiple times.

Through these strategies, the device garnered over 1.3 million pre-orders within seven hours of reservations opening on Huawei’s official e-commerce site, Vmall.com. By Monday afternoon, the Mate XT had received over 3 million pre-orders on Vmall, with Scalpers reselling it for at least 20,000 yuan ($2,821, £2,147) on the gray market, as reported by local media.

Official sales of the Mate XT are scheduled to commence on 20 September.

The 5G-capable Mate XT comes in red and black colors and offers 16GB of RAM, with internal storage options of either 512GB or 1TB, according to Huawei.

When folded, it resembles a standard smartphone; however, when unfolded, it reveals a large, nearly square screen, similar to a tablet.

Foldable smartphone shipments worldwide grew by 85 percent year-on-year in the April to June period, as per Tech Insights.

Huawei leads the global market due to its significant market share in China, followed by Samsung Electronics and China’s Vivo.

In the meantime, Apple has been promoting its AI plans since the beginning of this year, and these announcements have contributed to driving its stock price to record levels, reclaiming its position as the most valuable US-listed company ahead of Microsoft, Nvidia, and Google parent Alphabet.

Despite this, the company has encountered challenges in the crucial China smartphone market, with Huawei displacing it from the top five vendors in the quarter ending in July.

For the first time in history, all top five smartphone vendors in China in that quarter were domestic companies, researchers revealed.

Huawei’s success has been propelled by its Mate 60 flagship, introduced last summer, featuring a high-end chip produced domestically despite US sanctions on both Huawei and its chip-manufacturing partner SMIC.

Huawei has also been the leading seller of foldable smartphones in China for the past two quarters.

Honor, a Huawei spin-off, lags behind Huawei in China but secured the top position for foldable smartphones in Western Europe in the most recent quarter, according to Canalys.

It is now feasible to significantly expand the exploration of new materials.

It can require extended periods of meticulous effort — consulting information, conducting calculations, and performing precise laboratory tests — before scientists can introduce a new material with specific properties, whether it is for constructing an improved device or an enhanced battery. However, this process has been simplified thanks to artificial intelligence advancements.

A new AI algorithm named M3GNet, developed by researchers at the Jacobs School of Engineering at the University of California San Diego, is capable of predicting the structure and dynamic properties of any material, whether existing or newly conceptualized.

In fact, M3GNet was utilized to create a database containing over 31 million innovative materials that have not yet been synthesized, with their properties being forecasted by the machine learning algorithm. Moreover, this process occurs almost instantaneously.

Millions of potential materials

M3GNet can search for virtually any assigned material, such as metal, concrete, biological material, or any other type. To predict a material’s properties, the computer program needs to understand the material’s structure, which is based on the arrangement of its atoms.

In many ways, predicting new materials is similar to predicting protein structure — an area in which AlphaFold AI, developed by Google DeepMind, has excelled. Earlier this year, DeepMind announced the successful decoding of the structures of nearly all proteins in scientists’ catalogs, totaling over 200 million proteins.

Proteins, as the fundamental building blocks of life, undertake various functions within cells, from transmitting regulatory signals to protecting the body from bacteria and viruses. The accurate prediction of proteins’ 3D structures from their amino-acid sequences is therefore immensely beneficial to life sciences and medicine, and represents a revolutionary achievement.

Similar to how biologists previously struggled to decode only a few proteins over the course of a year due to inherent complexities, materials scientists can now invent and test novel materials much faster and more cost-effectively than ever before. These new materials and compounds can subsequently be integrated into batteries, drugs, and semiconductors.

“We need to know the structure of a material to predict its properties, similar to proteins,” explained Shyue Ping Ong, a nanoengineering professor at UC San Diego. “What we need is an AlphaFold for materials.”

Ong and his colleagues adopted a similar approach to AlphaFold, combining graph neural networks with many-body interactions to create a deep learning AI capable of scanning and generating practical combinations using all the elements of the periodic table. The model was trained using a vast database of thousands of materials, complete with data on energies, forces, and stresses for each material.

As a result, M3GNet analyzed numerous potential interatomic combinations to predict 31 million materials, over a million of which are expected to be stable. Additionally, the AI ​​​​can conduct dynamic and complex simulations to further validate property predictions.

“For instance, we are often interested in how fast lithium ions diffuse in a lithium-ion battery electrode or electrolyte. The faster the diffusion, the more quickly you can charge or discharge a battery,” Ong stated. “We have demonstrated that the M3GNet IAP can accurately predict the lithium conductivity of a material. We firmly believe that the M3GNet architecture is a transformative tool that significantly enhances our ability to explore new material chemistries and structures.”

The Python code for M3GNet has been made available as open-source on Github for those interested. There are already plans to integrate this powerful predictive tool into commercial materials simulation software.

Google AI discovers 2.2 million new materials for various technologies

DeepMind has propelled researchers centuries ahead of the traditional pace of materials discovery methods.

Inorganic crystals play a crucial role in modern technologies. Their highly-ordered atomic structures provide them with unique chemical, electronic, magnetic, or optical properties that are utilized in a wide range of applications, from batteries to solar panels, microchips to superconductors.

Creating innovative inorganic crystals in a laboratory — whether to enhance an existing technology or to power a new one — is theoretically simple. A researcher sets up the conditions, conducts the procedure, and learns from failures to adjust conditions for the next attempt. This process is repeated until a new, stable material is obtained.

However, in practice, the process is exceedingly time-consuming. Traditional methods rely on trial-and-error guesswork that either modifies a known crystalline structure or makes speculative attempts. This process can be costly, time-intensive, and, if unsuccessful, may leave researchers with limited insights into the reasons for failure.

The Materials Project, an open-access database established at the Lawrence Berkeley National Laboratory, has revealed that about 20,000 inorganic crystals were discovered through human experimentation. Over the past ten years, researchers have utilized computational techniques to increase that number to 48,000.

Google’s AI research lab, DeepMind, has unveiled the results of a new deep-learning AI designed to forecast the potential structures of previously unidentified inorganic crystals. And the outcomes are far ahead of schedule.

DeepMind’s latest AI, named the “Graph Networks for Materials Exploration” (GNoME), is a graph neural network that identifies connections between data points through graphs.

GNoME was trained using data from the Materials Project and began creating theoretical crystal structures based on the 48,000 previously found inorganic crystals. It made predictions using two pipelines.

The first pipeline, known as the “structural pipeline,” relied on known crystal structures for its predictions. The second pipeline, known as the “compositional pipeline,” took a more randomized approach to explore potential molecule combinations.

The AI ​​then validated its predictions using “density functional theory,” a method used in chemistry and physics to calculate atomic structures. Regardless of success or failure, this process generated more training data for the AI ​​to learn from, informing future pipeline predictions.

In essence, the AI’s pipelines and subsequent learning mimic the human experimental approach mentioned earlier, but with the advantage of the AI’s processing power for faster calculations.

The researchers emphasize that, unlike language or vision, in materials science, data can continue to be generated, leading to the discovery of stable crystals, which can be used to further expand the model.

Overall, GNoME forecasted 2.2 million new materials, with around 380,000 considered the most stable and potential candidates for synthesis.

The potential inorganic crystals include layered, graphene-like that compounds could aid in the development of advanced superconductors and lithium-ion conductors that may enhance battery performance.

The authors of the study, Google DeepMind researchers Amil Merchant and Ekin Dogus Cubuk, stated that “GNoME’s discovery of 2.2 million materials would be equivalent to about 800 years’ worth of knowledge and demonstrate an unprecedented scale and level of accuracy in predictions.”

The research findings were published in the peer-reviewed journal Nature, and the 380,000 most stable materials will be freely available to researchers through the Materials Project, thanks to DeepMind’s contribution.

GNoME’s predicted materials are theoretically stable, but only 736 of them have been experimentally verified to date. This suggests that the model’s predictions are accurate to some extent but also highlights the long road ahead for experimental fabrication, testing, and application of all 380,000 materials.

To bridge this gap, the Lawrence Berkeley National Laboratory assigned their new A-Lab, an experimental lab that combines AI and robotics for fully autonomous research, to synthesize 58 of the predicted materials.

A-Lab, a closed-loop system, can make decisions without human input and has the capability to handle 50 to 100 times as many samples per day as a typical human researcher.

Yan Zeng, a staff scientist at A-Lab, noted that “We’ve adapted to a research environment, where we never know the outcome until the material is produced. The whole setup is adaptive, so it can handle the changing research environment as opposed to always doing the same thing.”

During a 17-day run, A-Lab successfully synthesized 41 of the 58 target materials, at a rate of more than two materials per day and with a success rate of 71%. The NBNL researchers published their findings in another Nature paper.

The researchers are investigating why the remaining 17 inorganic crystals did not materialize. In some cases, this may be attributed to inaccuracies in GNoME’s predictions.

In some cases, expanding A-Lab’s decision-making and active-learning algorithms might produce more favorable outcomes. Two instances involved successful synthesis after human intervention was attempted.

As a result, GNoME has provided A-Labs and human-operated research facilities worldwide with ample material to work with in the foreseeable future.

“My goal with the Materials Project was not only to make the data I generated freely available to expedite materials design globally, but also to educate the world on the capabilities of computations.

They can efficiently and rapidly explore vast spaces for new compounds and properties, surpassing the capabilities of experiments alone,” stated Kristin Persson, founder and director of the Materials Project.

He further stated, “In order to tackle global environmental and climate challenges, we need to develop new materials. Through materials innovation, we could potentially create recyclable plastics, utilize waste energy, manufacture better batteries, and produce more durable and affordable solar panels, among other things.”

In November, DeepMind, Google’s AI division, released a press statement titled “Millions of New Materials Discovered with Deep Learning.” However, researchers who analyzed a subset of the discoveries reported that they had not yet come across any strikingly novel compounds in that subset .

“AI tool GNoME finds 2.2 million new crystals, including 380,000 stable materials that could power future technologies,” Google announced regarding the discovery, stating that this was “equivalent to nearly 800 years’ worth of knowledge,” many of the discoveries “defied previous human chemical intuition,” and it represented “a tenfold increase in stable materials known to humanity.” The findings were published in Nature and garnered widespread attention in the media as a legacy to the tremendous potential of AI in science.

Another paper, simultaneously released by researchers at Lawrence Berkeley National Laboratory “in collaboration with Google DeepMind, demonstrates how our AI predictions can be utilized for autonomous material synthesis,” Google reported.

In this experiment, researchers established an “autonomous laboratory” (A-Lab) that utilized “computations, historical data from the literature, machine learning, and active learning to plan and interpret the outcomes of experiments conducted using robotics.

” Essentially, the researchers employed AI and robots to eliminate human involvement in the laboratory, and after 17 days, they had discovered and synthesized new materials, which they asserted “demonstrates the effectiveness of artificial intelligence-driven platforms for autonomous materials discovery.”

However, in the past month, two external groups of researchers who analyzed the DeepMind and Berkeley papers and published their own DNA at the very least suggest that this specific research is being overhyped.

All the materials science experts I spoke to emphasized the potential of AI in discovering new types of materials. However, they contend that Google and its deep learning techniques have not made a groundbreaking advancement in the field of materials science.

In a perspective paper published in Chemical Materials this week, Anthony Cheetham and Ram Seshadri from the University of California, Santa Barbara selected a random sample of the 380,000 proposed structures released by DeepMind and stated that none of them met a three-part test to determine whether the proposed material is “credible,” “useful,” and “novel.”

They argue that what DeepMind discovered are “crystalline inorganic compounds and should be described as such, rather than using the more general term ‘material,’” which they believe should be reserved for substances that “demonstrate some utility.”

In their analysis, they wrote, “We have not yet come across any surprisingly novel compounds in the GNoME and Stable Structure listings, although we anticipated that there must be some among the 384,870 compositions.

We also note that, while many of the new compositions are minor adaptations of known materials, the computational approach produces reliable overall compositions, which gives us confidence that the underlying approach is sound.”

In a phone interview, Cheetham informed me, “The Google paper falls short in terms of offering a useful, practical contribution to experimental materials scientists.” Seshadri stated, “We believe that Google has missed the mark with this.”

“If I were seeking a new material to perform a specific function, I wouldn’t sift through over 2 million proposed compositions as suggested by Google,” Cheetham mentioned. “I don’t think that’s the most effective approach.

I think the general methodology probably works quite well, but it needs to be much more targeted towards specific needs, as none of us have enough time in our lives to evaluate 2.2 million possibilities and determine their potential usefulness.”

We dedicated a significant amount of time to examining a small portion of the proposals, and we discovered that not only were they lacking in functionality, but most of them, although potentially credible, were not particularly innovative as they were essentially variations of existing concepts.

According to a statement from Google DeepMind, they stand by all the claims made in the GNoME paper.

The GNoME research by Google DeepMind introduces a significantly larger number of potential materials than previously known to science, and several of the materials predicted have already been synthesized independently by scientists worldwide.

In comparison to other machine learning models, the open-access material property database, The Materials Project, has acknowledged Google’s GNoMe database as top-tier. Google stated that some of the criticisms in the Chemical Materials analysis, such as the fact that many of the new materials have known structures but use different elements, were intentional design choices by DeepMind.

The Berkeley paper asserted that an “autonomous laboratory” (referred to as “A-Lab”) utilized structures proposed by another project called the Materials Project and employed a robot to synthesize them without human intervention, yielding 43 “novel compounds.” While a DeepMind researcher is listed as an author on this paper, Google did not actively conduct the experiment.

Upon analysis, human researchers identified several issues with this finding. The authors, including Leslie Schoop of Princeton University and Robert Palgrave of University College London, pointed out four common shortcomings in the analysis and concluded that no new materials had been discovered in that work.

Each of the four researchers I spoke to express that while they believe an AI-guided process for discovering new materials holds promise, the specific papers they evaluated were not necessarily groundbreaking and should not be portrayed as such.

“In the DeepMind paper, there are numerous instances of predicted materials that are clearly nonsensical. Not only to experts in the field, but even high school students could identify that compounds like H2O11 (a DeepMind prediction) do not seem plausible,” Palgrave conveyed to me.

“There are many other examples of clearly incorrect compounds, and Cheetham/Seshadri provide a more diplomatic breakdown of this. To me, it seems that basic quality control has been neglected—for the machine learning to output such compounds as predictions is concerning and indicates that something has gone wrong.”

AI has been employed to inundate the internet with vast amounts of content that is challenging for humans to sift through, making it difficult to identify high-quality human-generated work.

While not a perfect comparison, the researchers I consulted with suggested that a similar scenario could unfold in materials science: Having extensive databases of potential structures does not necessarily facilitate the creation of something with a positive societal impact.

“There is some value in knowing millions of materials (if accurate), but how does one navigate this space to find useful materials to create?” Palgrave questioned. “It is better to have knowledge of a few new compounds with exceptionally useful properties than a million where you have no idea which ones are beneficial.”

Schoop pointed out that there are already “50k unique crystalline inorganic compounds, but we only understand the properties of a fraction of these. So, it is unclear why we need millions more compounds if we have not yet comprehended all the ones we do know. It might be more beneficial to predict material properties than simply new materials.”

While Google DeepMind maintains its stand on the paper and disputes these interpretations, it is fair to say that there is now considerable debate regarding the use of AI and machine learning for discovering new materials, the context, testing, and implementation of these discoveries, and whether inundating the world with massive databases of proposed structures will truly lead to tangible breakthroughs for society, or simply generate a lot of noise.

“We do not believe that there is a fundamental issue with AI,” Seshadri remarked. “We think it is a matter of how it is utilized. We are not traditionalists who believe that these techniques have no place in our field.”

AI tools and advanced robotics are accelerating the search for urgently needed new materials.

In the latest development, researchers at Google DeepMind reported that a new AI model has identified over 2.2 million hypothetical materials.

Out of the millions of structures predicted by the AI, 381,000 were identified as stable new materials, making them prime candidates for scientists to fabricate and test in a laboratory.
Current Situation: The advancement of novel materials is crucial for the development of the next generations of the electrical grid, computing, and other technologies such as batteries, solar cells, and semiconductor chips.

This has led to significant investments in materials science and engineering by countries worldwide, including the US, China, and India. According to a report released this week from Georgetown’s Center for Security and Emerging Technology, AI and materials science are the top recipients of US federal grants to industry over the past six years.

China currently leads the field of materials engineering in several key areas, including publications, employment, and degrees awarded in the field.
Background: Traditionally, new materials were discovered by modifying the molecular structure of existing stable materials to create a novel material.

This method predicted approximately 48,000 stable inorganic crystal structures, with more than half of them being discovered in the past decade. However, this process is expensive, time-consuming, and less likely to yield radically different structures as it builds on known materials.

DeepMind accomplished this feat by utilizing existing data from the Materials Project at Lawrence Berkeley National Laboratory (LBNL) and other databases to train the AI, which then expanded the dataset as it learned.

How it Works: DeepMind’s tool, known as Graph Networks for Materials Exploration (GNoME), employs two deep learning models that represent the atoms and bonds in a molecule as a graph.

One of the models starts with known crystal structures and substitutes elements to create candidate structures. The other model, aiming to go beyond known materials and generate more diverse materials, uses only the chemical formula or composition of a candidate material to predict its stability.

The pool of candidates is filtered, and the stability of each structure is determined through energy measurements.
The most promising structures undergo evaluation with quantum mechanics simulations, and the resulting data is fed back into the model in a training loop known as active learning.
The Intrigue: GNoME seemed to grasp certain aspects of quantum mechanics and made predictions about structures it had not encountered.

For instance, despite being trained on crystals consisting of up to four elements, the AI ​​system was able to discover five- and six-element materials, which have been challenging for human scientists, as stated by Ekin Dogus Cubuk, who leads the materials discovery team at DeepMind, during a press briefing.

The ability of an AI to generalize beyond its training data is significant. Keith Butler, a professor of computational chemistry materials at University College London, who was not involved in the research, emphasized the importance of exploring uncharted territories if these models are to be used for discovery.

What’s Next: Although predicting the stability of a potential structure does not guarantee its manufacturability.

In another paper published this week, researchers at LBNL shared the results from a lab equipped with AI-guided robotics for autonomous crystal synthesis. The material synthesis recipes were suggested by AI models that utilized natural language processing to analyze existing scientific papers and were then refined as the AI ​​system learned from its mistakes.

Over a span of 17 days, operating 24/7, the A-Lab successfully synthesized 41 out of 58 materials they attempted to create, averaging more than two materials per day. However, some routes of synthesis may be challenging or costly to automate due to involving intricate glassware, movement across a lab, or other complex steps, as pointed out by Butler.

The A-Lab’s failures in synthesizing 17 materials were attributed to factors such as the requirement for higher heating temperatures or the need for better material grinding, which are standard procedures in a lab but fall outside the current capabilities of AI.

Ultimately, a material’s performance in conducting heat, being electronically insulating, or fulfilling other functions is essential. However, the synthesis and testing of a material’s properties are costly and time-consuming aspects of the process of developing a new material. Moreover, similar to many other AI systems, the models do not provide explanations for their decision-making process, as highlighted by Butler.

Butler also emphasized the impact of competitions to predict new structures and the influence of large language models (LLM) such as ChatGPT and other generative AI in the field.

The University of Liverpool’s Materials Innovation Factory, the Acceleration Consortium at the University of Toronto, and other organizations are working on developing self-driving laboratories.
According to Olexandr Isayev, a chemistry professor at Carnegie Mellon University involved in their automated Cloud Lab, certain scientific experiments can be effectively automated with machine learning and AI, although not all of them.

He also mentioned that the future progress in the field of science will be driven by the combination of software and hardware.

The expansion of this open-access resource is crucial for scientists who are striving to create new materials for upcoming technologies. With the help of supercomputers and advanced simulations, researchers can avoid the time-consuming and often ineffective trial-and-error process that was previously necessary.

The Materials Project, an open-access database established at the Lawrence Berkeley National Laboratory (Berkeley Lab) of the Department of Energy in 2011, calculates the properties of both existing and predicted materials.

Researchers can concentrate on the development of promising materials for future technologies, such as lighter alloys to enhance car fuel efficiency, more effective solar cells for renewable energy, or faster transistors for the next generation of computers.

Google’s artificial intelligence laboratory, DeepMind, has contributed nearly 400,000 new compounds to the Materials Project, expanding the available information for researchers. This dataset includes details about the atomic arrangement (crystal structure) and the stability (formation energy) of materials.

The Materials Project has the capability to visually represent the atomic structure of various materials. For example, one of the new materials, Ba₆Nb₇O₂₁, was computed by GNoME and consists of barium (blue), niobium (white), and oxygen (green). This was acknowledged by the Materials Project at Berkeley Lab.

Kristin Persson, the founder and director of the Materials Project at Berkeley Lab and a professor at UC Berkeley, stated, “We need to develop new materials to address global environmental and climate challenges. Through innovation in materials, we have the potential to create recyclable plastics, utilize waste energy, produce better batteries, and manufacture more cost-effective solar panels with increased longevity, among other possibilities.”

The Role of GNoME in Material Discovery

Google DeepMind created a deep learning tool called Graph Networks for Materials Exploration, or GNoME, to generate new data. GNoME was trained using workflows and data that had been developed over a decade by the Materials Project, and the GNoME algorithm was refined through active learning.

Ultimately, researchers using GNoME generated 2.2 million crystal structures, including 380,000 that are being added to the Materials Project and are predicted to be stable, thus potentially valuable for future technologies. These recent findings from Google DeepMind were published in the journal Nature.

Some of the calculations from GNoME were utilized in conjunction with data from the Materials Project to test A-Lab, a facility at Berkeley Lab where artificial intelligence guides robots in creating new materials. The first paper from A-Lab, published in Nature, demonstrated that the autonomous lab can rapidly discover new materials with minimal human involvement.

During 17 days of independent operation, A-Lab successfully produced 41 new compounds out of 58 attempts – a rate of over two new materials per day. In comparison, it can take a human researcher months of trial and error to create a single new material , if they are able to achieve it at all.

To create the novel compounds forecasted by the Materials Project, A-Lab’s AI generated new formulas by analyzing scientific papers and using active learning to make adjustments. Data from the Materials Project and GNoME were utilized to assess the predicted stability of the materials.

Gerd Ceder, the principal investigator for A-Lab and a scientist at Berkeley Lab and UC Berkeley, stated, “We achieved an impressive 71% success rate, and we already have several methods to enhance it. We have demonstrated that combining theory and data with automation yields remarkable results. We can create and test materials more rapidly than ever before, and expanding the data points in the Materials Project allows us to make even more informed decisions.”

The Impact and Future of the Materials Project

The Materials Project stands as the most widely accessed open repository of information on inorganic materials globally. The database contains millions of properties on hundreds of thousands of structures and molecules, with data primarily processed at Berkeley Lab’s National Energy Research Science Computing Center.

Over 400,000 individuals are registered as users of the site, and on average, over four papers citing the Materials Project are published each day. The contribution from Google DeepMind represents the most significant addition of structure-stability data from a group since the inception of the Materials Project.

Ekin Dogus Cubuk, lead of Google DeepMind’s Materials Discovery team, expressed, “We anticipate that the GNoME project will propel research into inorganic crystals forward. External researchers have already verified over 736 of GNoME’s new materials through concurrent, independent physical experiments, demonstrating that our model’s discoveries can be realized in laboratories.”

This one-minute time-lapse illustrates how individuals across the globe utilize the Materials Project over a four-hour period. The data dashboard showcases a rolling one-hour window of worldwide Materials Project activity, encompassing the number of requests, the users’ country , and the most frequently queried material properties. Credit: Patrick Huck/Berkeley Lab

The Materials Project is currently processing the compounds from Google DeepMind and integrating them into the online database. The new data will be freely accessible to researchers and will also contribute to projects such as A-Lab that collaborate with the Materials Project.

“I’m thrilled that people are utilizing the work we’ve conducted to generate an unprecedented amount of materials information,” said Persson, who also serves as the director of Berkeley Lab’s Molecular Foundry. “This is precisely what I aimed to achieve with the Materials Project: not only to make the data I produced freely available to expedite materials design worldwide, but also to educate the world on the capabilities of computations. They can efficiently and explore rapidly large spaces for new compounds and properties more efficiently and rapidly than experiments alone.”

By pursuing promising leads from data in the Materials Project over the past decade, researchers have experimentally confirmed valuable properties in new materials across various domains. Some exhibit potential for use:

– in carbon capture (extracting carbon dioxide from the atmosphere)
– as photocatalysts (materials that accelerate chemical reactions in response to light and could be employed to break down pollutants or generate hydrogen)
– as thermoelectrics (materials that could aid in harnessing waste heat and converting it into electrical power)
– as transparent conductors (which could be beneficial in solar cells, touch screens, or LEDs)

However, identifying these potential materials is just one of many stages in addressing some of humanity’s significant technological challenges.

“Developing a material is not for the faint-hearted,” Persson remarked. “It takes a long time to transition a material from computation to commercialization. It must possess the right properties, function within devices, scale effectively, and offer the appropriate cost efficiency and performance. The objective of the Materials Project and facilities like A-Lab is to leverage data, enable data-driven exploration, and ultimately provide companies with more viable opportunities for success.”

The researchers at Google DeepMind claim they have increased the number of stable materials known by ten times. Some of these materials could have applications in various fields such as batteries and superconductors, if they can be successfully produced outside the laboratory.

The robotic chefs were busy in a crowded room filled with equipment, each one performing a specific task. One arm selected and mixed ingredients, another moved back and forth on a fixed track tending to the ovens, and a third carefully plated the dishes.

Gerbrand Ceder, a materials scientist at Lawrence Berkeley National Lab and UC Berkeley, heard in approval as a robotic arm delicately closed an empty plastic vial—a task that he particularly enjoyed observing. “These robots can work tirelessly all night,” Ceder remarked, giving two of his graduate students a wry smile.

Equipped with materials like nickel oxide and lithium carbonate, the A-Lab facility is designed to create new and intriguing materials, particularly those that could be valuable for future battery designs. The outcomes of the experiments can be unpredictable.

Even a human scientist often makes mistakes when trying out a new recipe for the first time. Similarly, the robots sometimes produce a fine powder, while other times the result is a melted sticky mess or everything evaporates, leaving nothing behind. “At that point , humans would have to decide what to do next,” Ceder explained.

The robots are programmed to do the same. They analyze the results of their experiments, adjust the recipe, and try again, and again, and again. “You give them some recipes in the morning, and when you come back home, you might find a beautifully made soufflé,” said materials scientist Kristin Persson, Ceder’s close collaborator at LBNL (and also his spouse). Or you might return to find a burnt mess. “But at least tomorrow, they will make a much better soufflé.”

Recently, the variety of materials available for Ceder’s robots has expanded significantly, thanks to an AI program developed by Google DeepMind. This software, called GNoME, was trained using data from the Materials Project, a freely accessible database of 150,000 known materials overseen by Persson .

Using this information, the AI ​​system generated designs for 2.2 million new crystals, out of which 380,000 were predicted to be stable—unlikely to decompose or explode, making them the most feasible candidates for synthesis in a lab. This has expanded the range of known stable materials almost tenfold. In a paper published in Nature, the authors state that the next solid-state electrolyte, solar cell materials, or high-temperature superconductor could potentially be found within this expanded database.

The process of discovering these valuable materials starts with actually synthesizing them, which emphasizes the need to work quickly and through the night. In a recent series of experiments at LBNL, also published in Nature, Ceder’s autonomous lab successfully created 41 of the theorized materials over 17 days, helping to validate both the AI ​​​model and the lab’s robotic techniques.

When determining if a material can be synthesized, whether by human hands or robot arms, one of the initial questions to ask is whether it is stable. Typically, this means that its atoms are arranged in the lowest possible energy state. Otherwise, the crystal will naturally want to transform into something else.

For thousands of years, people have been steadily adding to the list of stable materials, initially by observing those found in nature or discovering them through basic chemical intuition or accidents. More recently, candidate materials have been designed using computers.

According to Persson, the issue lies in bias: Over time, the collective knowledge has come to favor certain familiar structures and elements. Materials scientists refer to this as the “Edison effect,” based on Thomas Edison’s rapid trial-and-error approach to finding a lightbulb filament, testing thousands of types of carbon before settling on a variety derived from bamboo.

It took another decade for a Hungarian group to develop tungsten. “He was limited by his knowledge,” Persson explained. “He was biased, he was convinced.”

The approach by DeepMind aims to overcome these biases. The team started with 69,000 materials from Persson’s database, which is freely accessible and supported by the US Department of Energy. This was a good starting point, as the database contains detailed energetic information necessary to understand why certain materials are stable and others are not.

However, this was not enough data to address what Google DeepMind researcher Ekin Dogus Cubuk calls a “philosophical contradiction” between machine learning and empirical science. Similar to Edison, AI struggles to generate genuinely new ideas beyond what it has already encountered.

“In physics, you never want to learn something you already know,” Cubuk stated. “You almost always want to make generalizations outside of the known domain”—whether that involves discovering a different class of battery material or a new theory of superconductivity.

GNoME uses active learning, where a graph neural network (GNN) nests the database to identify patterns in stable structures and minimize atomic bond energy in new structures across the periodic table. This process generates numerous potentially stable candidates.

These candidates are then verified and adjusted using density-functional theory (DFT), a quantum mechanics technique. The refined results are incorporated back into the training data, and the process is repeated.

Through multiple iterations, the approach was able to produce more complex structures than those in the original Materials Project data set, including some composed of five or six unique elements, surpassing the four-element cap of the training data.

However, DFT is only a theoretical validation, and the next step involves the actual synthesis of materials. Ceder’s team selected 58 crystals to create in the A-Lab, considering the lab’s capabilities and available precursors. The initial attempts failed, but after multiple adjustments , the A-Lab successfully produced 41 of the materials, or 71 percent.

Taylor Sparks, a materials scientist at the University of Utah, notes the potential of automation in materials synthesis but emphasizes the impracticality of using AI to propose thousands of new hypothetical materials and then pursuing them with automation. He stresses the importance of tailoring efforts to produce materials with specific properties rather than blindly generating a large number of materials.

While GNNs are increasingly used to generate new material ideas, there are concerns about the scalability of the synthesis approach. Sparks mentions that the challenge lies in whether the scaled synthesis matches the scale of the predictions, which he believes is currently far from reality.

Only a fraction of the 380,000 materials in the Nature paper are likely to be feasible for practical synthesis. Some involve radioactive or prohibitively expensive elements, while others require synthesis under extreme conditions that cannot be replicated in a lab.

This practicality challenge extends to materials with potential for applications such as photovoltaic cells or batteries. According to Persson, the bottleneck consistently lies in the production and testing of these materials, especially for those that have never been made before.

Furthermore, turning a basic crystal into a functional product is a lengthy process. For instance, predicting the energy and structure of a crystal can help understand the movement of lithium ions in an electrolyte, a critical aspect of battery performance. However, predicting the electrolyte’s interactions with other materials or its overall impact on the device is more challenging.

Despite these challenges, the expanded range of materials opens up new possibilities for synthesis and provides valuable data for future AI programs, according to Anatole von Lilienfeld, a materials scientist at the University of Toronto.

Additionally, the new materials generated by GNoME have piqued the interest of Google. Pushmeet Kohli, vice president of research at Google DeepMind, likes GNoME to AlphaFold and emphasizes the potential for exploring and expanding the new data to address fundamental problems and accelerate synthesis using AI .

Kohli stated that the company is considering different approaches to directly engage with physical materials, such as collaborating with external labs or maintaining academic partnerships. He also mentioned the possibility of establishing its own laboratory, referring to Isomorphic Labs, a spinoff from DeepMind focused on drug discovery, which was founded in 2021 after the success of AlphaFold.

Researchers may encounter challenges when attempting to apply the materials in practical settings. The Materials Project is popular among both academic institutions and businesses because it permits various types of use, including commercial activities.

Google DeepMind’s materials are made available under a distinct license that prohibits commercial usage. Kohli clarified, “It’s released for academic purposes. If individuals are interested in exploring commercial partnerships, we will evaluate them on a case-by-case basis.”

Several scientists working with new materials observed that it’s unclear what level of control the company would have if experimentation in an academic lab leads to a potential commercial application for a GNoME-generated material. Generating an idea for a new crystal without a specific purpose in mind is generally not eligible for a patent, and it could be challenging to trace its origin back to the database.

Kohli also mentioned that although the data is being released, there are currently no plans to release the GNoME model. He cited safety concerns, stating that the software could potentially be used to generate hazardous materials, and expressed uncertainty about Google DeepMind’s materials strategy. ” It’s difficult to predict the commercial impact,” Kohli said.

Sparks anticipates that his colleagues in academia will be displeased with the absence of code for GNoME, similar to biologists’ reaction when AlphaFold was initially published without a complete model. “That’s disappointing,” he remarked.

Other materials scientists are likely to want to replicate the findings and explore ways to enhance the model or customize it for specific purposes. However, without access to the model, they are unable to do so, according to Sparks.

In the interim, the researchers at Google DeepMind hope that the discovery of hundreds of thousands of new materials will be sufficient to keep both human and robotic theorists and synthesizers occupied. “Every technology could benefit from improved materials. It’s a bottleneck,” Cubuk remarked “This is why we need to facilitate the field by discovering more materials and assisting people in discovering even more.”