Donald Trump’s vice-presidential pick has indicated he favors a hands-off approach to A.I. but also wants to increase scrutiny of the biggest tech firms in the field.

Mr. Vance spent less than five years in Silicon Valley’s tech industry, but the connections he made with Peter Thiel and others became crucial to his political ascent.

Windows 11’s much-awaited 24H2 update might arrive a bit later than was previously rumored, or at least that’s the most recent chatter and speculation online.

Windows Latest picked up on a Microsoft update regarding the potential release timeframe for 24H2 (as part of the introduction of smaller updates for Windows 11 that we just reported on).

Microsoft tells us that: “Windows 11, version 24H2 will be available as a traditional feature update to all devices later this year.”

Of course, that doesn’t mention any specific timeframe, but as Windows Latest and others have theorized, the wording that Microsoft employs here, “later this year,” is a hint (admittedly a vague one) that it’ll be Q4 rather than Q3.

Okay, so technically, later this year could mean next month, but if we were looking at an August 2024 release, it’d be a strange choice of phrasing, let’s face it.

The current rumor for the debut of Windows 11 24H2 is September 2024, but perhaps given this small clue dropped by Microsoft, we might be waiting a little longer. There are other reasons to believe this may be the case, so let’s dive into those next.

(Image credit: Future / John Loeffler) Analysis: A case of spinning too many plates?

The reality is that Microsoft is doing a very different kind of update this year, with the annual upgrade for Windows 11 already having been made available – on Copilot+ PCs, anyway. That said, this isn’t the full truth of the matter, as what’s installed on those AI-focused laptops is a barebones version of Windows 11 24H2 – which is required because Copilot+ devices need to have the underpinning platform brought in with 24H2 (called Germanium).

That installation of 24H2 on Copilot+ PCs does not have all the features that’ll be introduced with the full 24H2 update when it arrives later this year. Everyone on Windows 11, Copilot+ and normal PCs alike, will get that update, with all the features on-board, simultaneously rolled out as normal.

The thing is, though, that Microsoft has had a lot more on its plate with the annual update process in 2024, due to having to do all the groundwork for Copilot+ PCs (and the work behind the scenes on Germanium). Heap on top of that a bunch of effort needed to realize new AI features, and some big ones too, including Recall, the controversial addition that was pulled. Indeed, Recall is a flagship AI feature, and one Microsoft is currently having to completely revamp (it’s even been yanked from test builds for now). Again, that’s more work.

So, this is why we aren’t particularly optimistic about a September 2024 release date, as previously rumored, any longer. Simply put, Microsoft is spinning so many plates with preparing 24H2, and other AI features, plus dealing with an almighty headache around Recall, that it’s all too easy to believe a Q4 release is at least somewhat likely for 24H2. After all, it could land as late as November (which has happened in the past with an H2 update for Windows), and Microsoft might just need more breathing room this time around.

This is, of course, pure speculation, and 24H2 could make the cut for Q3 – or September, rather (we’d be seriously shocked at an August release, frankly). As ever, we’ll wait and see, but following the progress of 24H2 in testing should give us more clues as 2024 rolls onwards.

You might also like...

• Windows 11 users are getting fed up with the performance of the OS – including an ex-Microsoft senior engineer
• Don’t make these 5 big mistakes when using Windows 11
• You can run Windows 11 seamlessly on Apple silicon Macs

The Meta Quest 3 Netflix app has been officially deprecated – if you try and use it to stream content to your Quest VR headset it’ll no longer work, and you’ll get a playback error instead. But the new browser-based experience that replaces it is leaps and bounds ahead of what we had before, so you won’t miss the old streaming method.

The Netflix app was one of the few streaming apps you could download on Quest, but it was flawed (to put it politely). There was no passthrough support to watch in mixed reality, and streams were capped at 480p (standard definition, or SD).

While the app had its supporters they were few and far between, and the Netflix app was deemed by many, us included, as a waste of digital library space. So it came as no surprise that Netflix would come up with a different strategy for supporting the Quest 3 – ditching the app completely and instead bringing its streaming service to the Quest browser.

We’ve tested it out and the difference is night-and-day. Streams are now a significantly crisper 1080p (full-HD), and you can choose to watch Netflix content in mixed reality or in your virtual home environment. What’s more you can move the window around to position it for comfortability or multitasking, you can view on a curved or flat screen and resize the display as required, and you can dim or brighten the environment to make it feel more or less like you’re watching in your own private VR movie theater.

It’s turned Netflix-in-VR from a niche oddity to a genuinely great example of the tech. Now we just need other streaming platforms to follow suit.

The first step for VR streaming

It’s worth noting you can use the browser to watch content from other best streaming services – like Disney Plus – but when testing these other platforms there was a noticeable dip in quality compared to Netflix. 

Apple's Vision Pro is much better for VR video (Image credit: Apple)

The darkened cinema-like experience remains (as that’s a feature of the browser itself) but the picture quality is grainier, perhaps due to the services not offering the same level of browser support as Netflix. 

What’s more, while browser-based experiences are generally fine it’s not a perfect solution, even on the Netflix side of things. Primarily because you can’t download shows or films to watch offline – say, when you take your next flight. This was lacking in the original Netflix VR app too, but now it’s a feature that might be impossible to implement unless the platform creates a new dedicated app.

Apps for Disney Plus and Prime Video on the Apple Vision Pro do support downloads, so perhaps they’ll keep this functionality when they get a Quest 3 port (although who knows if that’ll happen). 

Meta hasn't made video content as much of a focus for Quest as Apple has with the Vision Pro, but if the Netflix experience proves popular we could see it invest more in that aspect of VR. We'll just have to watch this space.

You might also like

• Even in VR, Taskmaster is way harder than I thought it would be
• Apple Vision Pro finally hits the UK – here's how you can try one
• Vision Pro at one – I love Apple's revolutionary headset, so why do I hardly ever use it?

AI brand Anthropic is bringing its Claude app to Android two months after making its debut on iOS. Claude, in case you’re not familiar with it or need a refresher, is a generative AI platform similar to ChatGPT. It answers questions given to it via prompts. These prompts can include paragraphs of text, files, images, or a combination of all three at once. 

As revealed in a recent announcement post, the Android app runs on Claude 3.5 Sonnet which the company states is their most powerful model yet. It possesses multilingual processing to provide “real-time language translation” and “advanced reasoning”. The latter allows the AI to help with complex tasks such as helping users analyze data for work.

Plus, just like the iOS app, a conversation with Claude can continue off your smartphone and pick up on the browser version or an iPhone. All you need to do is make sure the account you’re using is the same across devices.

Warnings and limitations

Claude is available for download off the Google Play Store. 

Upon opening the app for the first time, you’ll be asked to sign in or create an account if you don’t have one already. You’ll also receive a notice informing users that the AI isn’t perfect. Anthropic warns people “Claude may occasionally generate incorrect or misleading information”. The AI is not meant to give advice on anything so don’t rely on just the conversation without doing your own research. 

The Claude app is easy to use and sports a clean UI, free of distractions. Tapping the paperclip icon in the bottom left corner lets you add multimodal content to the chat. Past interactions appear on the starting page.

(Image credit: Future)

Be aware that the free app has a daily limit on the amount of messages you can receive. This limit can vary. According to Anthropic, it all depends on how much demand there is on a given day and they’ll notify you once you’ve reached the cap.

Of course, users can always purchase a Claude Pro subscription to increase the daily limit up to five times the amount people on the free plan get and gain extra features such as access to Anthropic’s other models.

Emphasis on privacy

Anthropic certainly has its work cut out for it if the company aims to take on rivals OpenAI and Google. It cannot be understated just how dominant ChatGPT is in the industry. TechCrunch points out that Claude on iOS saw a “tepid reception during its launch” receiving 157,000 total global downloads in its first week. ChatGPT saw 480,000 installs during its first five days.

Claude does have advantages that set it apart. Its developer is more committed to ensuring user privacy. Anthropic states it does not collect people’s inputs or outputs to utilize in training their AI unless they’re given explicit consent to do so. What’s more, the AI has been shown to be quite good at solving complex reasoning problems and writing nuanced answers

So if you want a chatbot on Android with an emphasis on informative responses and privacy, then Claude AI comes highly recommended. For those wanting other alternatives, check out TechRadar’s round-up of the top seven chatbots you can try for free. 

You might also like

• You and your friends can now share and remix your favorite conversations with the Claude AI chatbot
• Survey says most believe generative AI is conscious, which may prove it's good at making us hallucinate, too
• Anthropic wants its AI assistant Claude to be your favorite coworker

The social media and rocket companies are based in California, which the billionaire criticized for its recent transgender legislation.

Artificial intelligence models require as much useful data as possible to perform but some of the biggest AI developers are relying partly on transcribed YouTube videos without permission from the creators in violation of YouTube's own rules, as discovered in an investigation by Proof News and Wired. 

The two outlets revealed that Apple, Nvidia, Anthropic, and other major AI firms have trained their models with a dataset called YouTube Subtitles incorporating transcripts from nearly 175,000 videos across 48,000 channels, all without the video creators knowing.

The YouTube Subtitles dataset comprises the text of video subtitles, often with translations into multiple languages. The dataset was built by EleutherAI, which described the dataset's goal as lowering barriers to AI development for those outside big tech companies. It's only one component of the much larger EleutherAI dataset called the Pile. Along with the YouTube transcripts, the Pile has Wikipedia articles, speeches from the European Parliament, and, according to the report, even emails from Enron. 

However, the Pile has a lot of fans among the major tech companies. For instance, Apple employed the Pile to train its OpenELM AI model, while the Salesforce AI model released two years ago trained with the Pile and has since been downloaded more than 86,000 times.

The YouTube Subtitles dataset encompasses a range of popular channels across news, education, and entertainment. That includes content from major YouTube stars like MrBeast and Marques Brownlee. All of them have had their videos used to train AI models. Proof News set up a search tool that will search through the collection to see if any particular video or channel is in the mix. There are even a few TechRadar videos in the collection, as seen below.

(Image credit: Proof News) Secret Sharing

The YouTube Subtitles dataset seems to contradict YouTube’s terms of service, which explicitly fobird automated scraping of its videos and associated data. That’s exactly what the dataset relied on, however, with a script downloading subtitles through YouTube’s API. The investigation reported that the automated download culled the videos with nearly 500 search terms. 

The discovery provoked a lot of surprise and anger from the YouTube creators Proof and Wired interviewed. The concerns about the unauthorized use of content are valid, and some of the creators were upset at the idea their work would be used without payment or permission in AI models. That’s especially true for those who found out the dataset includes transcripts of deleted videos, and in one case, the data comes from a creator who has since removed their entire online presence.

The report didn’t have any comment from EleutherAI. It did point out that the organization describes its mission as democratizing access to AI technologies by releasing trained models. That may conflict with the interests of content creators and platforms, if this dataset is anything to go by. Legal and regulatory battles over AI were already complex. This kind of revelation will likely make the ethical and legal landscape of AI development more treacherous. It’s easy to suggest a balance between innovation and ethical responsibility for AI, but producing it will be a lot harder. 

You might also like

• YouTube will now take down AI deepfakes of you if you ask
• Want to skip to the good bit of a video? YouTube is testing a smart AI feature for that
• YouTube will use AI to snip copyrighted music and not silence your whole video

The tech community, led by Sam Altman of OpenAI, has funded programs that give people unconditional cash. Some say it’s time to scale up.

The Xreal Beam Pro delivers much of what spatial computing has been missing on the AR glasses side of things. 

It finally turns your Xreal Air or Xreal Air 2 specs into a complete XR package rather than just a wearable display. The handheld, smartphone-like device adds much-needed features like easy spatial photo capture without an iPhone 15 Pro and helpful design details like two USB-C ports. Plus, it’s super affordable at only $199 / £189 for its base model from Xreal's official store.

Unfortunately, it falls short of solving every issue facing AR smart glasses – simultaneously throwing in a few annoyances that are exclusive to this currently imperfect product.

There’s a lot to love about the Xreal Beam Pro – for people who use their Xreal glasses frequently it’s not far from being a must-buy accessory – but it falls short of transforming AR smart glasses into the mainstream spatial computing gadget like I hoped it might.

Almost everything I wanted

Starting with what makes the Beam Pro fantastic, the spatial image-enabled cameras, which sit on the back of the device, are an XR convenience I didn’t know I needed. This dual-camera rig can shoot 1080p 3D video at 60FPS or snap a 50MP 3D photo, and the results look fantastic. 

As our own Lance Ulanoff has commented when viewing spatial images with the Vision Pro they’re a cut above flat snaps, and I wholeheartedly agree. So much so that I now carry the Beam Pro with me even without my Xreal glasses so I can capture memories in their full 3D glory.

(Image credit: Xreal)

The other huge convenience is the second USB-C port. Wired AR Glasses can rinse through your connected device’s battery – it’s a problem I’ve had when using every brand’s wired specs with a smartphone – so the ability to charge your Beam Pro while using it is a benefit that can’t go understated.

Beyond this, the Beam Pro’s strengths are simplicity and price. Thanks to hardware oddities XR glasses aren’t compatible with every product you’d think they would be, and buying adapters (if there are any) can be confusing even for the tech-savvy. The Beam Pro cuts through this chaos as the one add-on you need to complete your XR setup – with connecting your AR glasses to other tech now relegated to being an added bonus rather than a necessity. Plus at $199 / £189, the Xreal Beam Pro is a heck of a lot cheaper than the phones it’s imitating, with a lot to offer.

That said it does suffer technically at times.

Odd sacrifices were made

The base model (which I tested) comes with 128GB of storage which is fine, though the 6GB of RAM is disappointing and can lead to some sluggishness; opting for the 256GB upgraded model which comes with 8GB of RAM for just $249 / £239 looks to me like a more than worthwhile upgrade.

The display seems to struggle too, perhaps from a lack of processing power. When using the following mode – where the screen moves so that it’s always in front of you – the image flickers every so often and noticeably lags behind my head movements. Meanwhile, the body-anchor mode – which should keep the screen in one stationary place, like using a virtual TV – doesn’t always work as intended as the screen will steadily drift from its set position. Saying that, things did improve in this department thanks to software updates in the past day or two suggesting it could also be that the Nebula OS needs some refinements.

At its low price sacrifices should be expected, but when you consider a complete Xreal Air 2 glasses and Beam Pro package would set you back at least $598 / £588 (more than a Meta Quest 3), I'd like something with more oomph – at least 4K visuals, and 8GB of RAM as a starting point.

(Image credit: Future)

The software offers its own annoyances. While the VisionOS-like app layout is aesthetically fine, the app management is frustrating – early smartphone days levels of frustrating. 

The apps in this virtual layout don’t appear to have any particular order, and trying to rearrange the chaos by dragging icons around is finicky to say the least. The option to easily disable apps for this XR menu – so you aren’t just seeing everything installed on your Beam Pro – would be a start to make it less unwieldy. The next steps should then be easier icon rearrangement tools (even an on Beam Pro option so you can use touch controls to drag apps around) and the ability to make grouped folders of apps – but it’s yet to be seen how extensive and frequent Xreal’s OS updates will be.

To round off my gripes, the Xreal Beam Pro is too big. If this was a phone the additional screen real estate would be a blessing, not a curse, but the Beam Pro (despite its looks) is not a phone. Its goal is to be a portable accessory and controller for your Xreal glasses, and thanks to its 6.5-inch screen it struggles a tad in this department – it noticeably juts out of my jeans’ back pocket, and when using the controller I have to constantly readjust my grip (or use two hands) to reach every icon.

The Xreal Beam Pro is not a standalone product, and so it doesn’t need features like a large LCD 2K display when it could get away with something smaller with lower fidelity. I’d go as far as to say this thing doesn’t even need speakers – I’d happily lose them for more RAM or simply a lower price.

Is the Xreal Beam Pro right for you? 

(Image credit: Xreal)

If you’re looking for a portable smart TV for your AR glasses, and spatial photo capture device at an affordable price, then the Xreal Beam Pro won’t disappoint. In this regard, the Xreal beam Pro is close to perfect.

But as a spatial computer it’s not powerful enough nor refined enough (at least not yet) to be a standout winner. I’m excited to see what improvements Xreal makes to the Beam Pro’s software in the coming months, and to see what hardware comes next

You might also like

• Meta can't stop leaking its next VR headset, as it accidentally shows off the Quest 3S
• Meta has wild plans for camera-equipped headphones
• Meta’s massive OS announcement is more exciting than a Meta Quest 4 reveal

 When you interact with ChatGPT and other conversational generative AI tools, they process your input through algorithms to compose a response that can feel like it came from a fellow sentient being despite the reality of how large language models (LLMs) function. Two-thirds of those surveyed for a study by the University of Waterloo nonetheless believe AI chatbots to be conscious in some form, passing the Turing Test of convincing them that an AI is equivalent to a human in consciousness. 

Generative AI, as embodied by OpenAI’s work on ChatGPT, has progressed by leaps and bounds in recent years. The company and its rivals often talk about a vision for artificial general intelligence (AGI) with human-like intelligence. OpenAI even has a new scale to measure how close their models are to achieving AGI. But, even the most optimistic experts don’t suggest that AGI systems will be self-aware or capable of true emotions. Still, of the 300 people participating in the study, 67% said they believed ChatGPT could reason, feel, and be aware of its existence in some way. 

There was also a notable correlation between how often someone uses AI tools and how likely they are to perceive consciousness within them. That’s a testament to how good ChatGPT is at mimicking humans, but it doesn’t mean the AI has awakened. The conversational approach of ChatGPT likely makes them seem more human even, though no AI model works like a human brain at all. And while OpenAI is working on an AI model capable of doing research autonomously called Strawberry, that’s still different from an AI that is aware of what it is doing and why.

“While most experts deny that current AI could be conscious, our research shows that for most of the general public, AI consciousness is already a reality,” University of Waterloo professor of psychology and co-lead of the study Dr. Clara Colombatto explained. “These results demonstrate the power of language because a conversation alone can lead us to think that an agent that looks and works very differently from us can have a mind.”

(Image credit: Neuroscience of) Customer Disservice

The belief in AI consciousness could have major implications for how people interact with AI tools. On the positive side, it encourages manners and makes it easier to trust what the tools do, which could make them easier to integrate into daily life. But trust comes with risk, from overreliance on them for decision-making to, at the extreme end, emotional dependence on AI and fewer human interactions.

The researchers plan to look deeper into the specific factors making people think that AI has consciousness and what that means on an individual and societal level. It will also include long-term looks at how those attitudes change over time and with regard to cultural background. Understanding public perceptions of AI consciousness is crucial not only to developing AI products but also to the regulations and rules governing their use.  

“Alongside emotions, consciousness is related to intellectual abilities that are essential for moral responsibility: the capacity to formulate plans, act intentionally, and have self-control are tenets of our ethical and legal systems,” Colombatto said. “These public attitudes should thus be a key consideration in designing and regulating AI for safe use, alongside expert consensus.”

You Might Also Like

• OpenAI has a new scale for measuring how smart their AI models are becoming – which is not as comforting as it should be
• Survey finds most people would rather switch companies than deal with AI customer service
• OpenAI may be working on AI that can perform research without human help – which should go fine

It’s no secret that tracking is pretty commonplace on the web, whether on a desktop or when quickly searching on the phone. To highlight both the privacy concerns and a browser that aims to squash those trackers – or rather cause flying security cameras to explode – Apple’s rolling out a new campaign touting Safari’s privacy chops. 

The campaign, which debuts today as a short-form film and placements designed for desktop and mobile, aims to highlight the benefits of using Safari versus Google Chrome or another browser. It follows everyday folks as they browse the web for a whole host of reasons, with flying security cameras following them around and, honestly, spooking them. 

Apple’s aim is to remind folks of the privacy protections built into Safari – it was the first web browser to debut a “Private Browsing” experience in 2005, and since then that has been bolstered and is now locked behind Face ID, Touch ID, or a Passcode. Safari also blocks cross-site tracking, strips unnecessary trackers from links, and shades location data. 

Going a step further, Apple deploys machine learning to determine which sites might be tracking you under a feature dubbed “Intelligent Tracking Prevention.” When it spots a site tracking you, generally in the same vein as a cookie, Safari stops it and removes the tracker. 

You can see it working in real-time in the Safari Privacy Report on your device, which lists the sites by the number of trackers, the most common ones, and how many have been blocked or stopped.

Of course, Safari also masks your IP address, and when a site requests more precise location information, it asks you, the end user, to approve or deny it. All of this is part of Apple's four main pillars of privacy within Safari. 

The advertisements themselves are pretty compelling; on mobile, it’ll be a vertical takeover with someone peeling back the webpage from the top, and the tagline “Your browsing is being watched.” appears with “Safari stops it” shortly after. The idea, of course, is to get more folks using these privacy features if they’re already on Safari – and remember, most of these settings are on by default – or likely to get folks to use Safari on a Mac, iPhone, iPad, or Vision Pro. 

The full-length film paints a none-too-subtle picture of how Apple views the privacy differences between Android phones and the iPhone. It shows security cameras lurking about them while the iPhone stops the cameras in their tracks and causes them to explode.

Still, it's one of Apple's more jovial films, highlighting a pretty serious feature set baked into web browsing on its devices.

You Might Also Like

• Amazon Prime Day: the 99 best deals
• Apple secretly gave the iPhone's flashlight a major upgrade with iOS ...
• I lost my Kindle on an airplane – this is how I'll choose my new one ...

Want smaller updates for Windows 11? Naturally, everyone does – the smaller the better – and there’s some good news on this front, namely that most updates for the OS will be a lot more manageable, download-wise, soon enough.

This is because Microsoft is introducing a system of ‘checkpoint cumulative updates’ starting from the release of Windows 11 24H2 later this year (and indeed with Windows Server 2025, too).

Cumulative updates are the main monthly updates delivered to your Windows 11 PC (on the second Tuesday of every month), and the idea here is that much smaller updates are applied on top of milestone ‘checkpoint’ updates, ones that only make tweaks to that current checkpoint build.

As Microsoft explains in a blog post: “This will allow you to get features and security enhancements via the latest cumulative update through smaller, incremental differentials containing only the changes since the previous checkpoint cumulative update. This means that you can save time, bandwidth, and hard drive space.”

It’s a very useful step to take with streamlining updates, and we’re likely to see multiple smaller updates coming through, in between the checkpoint cumulative updates that’ll pitch up more rarely (these will be normal-sized downloads).

Analysis: Sterling work from Microsoft

What’s really happening, as Ghacks, who spotted this, points out, is that you can think of a ‘checkpoint cumulative update’ as a new base patch version for Windows 11. So, a bit like an entirely new version (such as 23H2, 24H2), but only for recently applied security fixes and other patches. Normally, these would all be bundled up in every single cumulative update, but now, as the checkpoint updates carry them – providing a new base version – there’s no longer any need to do so (hence the between-checkpoint updates get much trimmer).

All this will happen automatically via Windows Update as usual, and you won’t notice any difference or have to do anything. As noted, the only difference will be most updates (non-checkpoint ones) will be considerably quicker to download and install, and you’ll save a bit of storage space (perhaps an even more important benefit for those with, say, a smaller SSD in an affordable laptop).

When Windows 11 first launched, Microsoft worked to streamline its cumulative updates by using new compression technology. That was a sterling effort, too, and eventually the fruits of that labor – which cut update size by 40% – came to Windows 10 users. Whether this latest change will eventually benefit those on Windows 10 is more doubtful, though, as with its End of Life only just over a year away, Microsoft may not think it’s worthwhile to apply this move to the older operating system.

You might also like...

• Microsoft’s embarrassment over Recall fiasco gets worse as Windows 11 feature becomes the butt of Apple exec’s joke
• Don’t make these 5 big mistakes when using Windows 11
• Windows 11 or bust: Microsoft is insistently urging Windows 10 users to move on, or get left behind

Windows 11 is getting some work done on its interface that might consist of more minor tweaks, but nonetheless useful ones, and handy time-savers in some cases.

All this is packaged up in a new beta build (preview version 22635) recently released by Microsoft, and the first noteworthy change is to File Explorer.

Specifically, it’s to the Home page in File Explorer, which now has an option to list all the files recently shared with you. So, if someone has shared a document on Teams, or via email – or indeed other avenues – all the most recently shared material will be listed (chronologically), along with key details such as who shared it.

Another nifty change is inbound for the taskbar, and this one is a streamlining measure. Essentially, in the system tray, on the far right of the taskbar, Microsoft has shortened the time and date so it takes up less space (and doesn’t show the year), plus the notifications bell has been dropped (though you can turn it back on if you want).

On top of this, Windows Studio Effects is now available to access via Quick Settings in the system tray, allowing you easy access to the feature – which provides a bunch of AI-driven camera and audio tricks for PCs that have an NPU (for accelerating AI workloads).

What’s neat here is that if you’re using an app that can leverage Windows Studio Effects – which offers abilities like a background blur feature, or filtering out background noise so you can better hear the speaker – an icon will pop up in the system tray to allow you to directly launch Studio Effects settings in Windows 11.

There are various other minor tweaks with build 22635 and Microsoft’s blog post explains them all at length.

Analysis: Handiness abounds

A trimmer system tray is welcome, certainly, as is the useful touch of being able to access Windows Studio Effects settings much more easily (and to know when an app can utilize these).

Furthermore, the change to File Explorer is another handy move, allowing you to quickly and conveniently access shared material. If you know someone has shared something recently, but can’t find the file, you can just pop open File Explorer on the desktop and look in this list. A useful shortcut indeed, potentially, though one that’ll eventually be overshadowed by the Recall feature, in theory, for Copilot+ PCs – when it’s relaunched. (For folks who want to use Recall, anyway, and we’re guessing after all the controversy, not everyone will).

It’s worth noting that build 22635 comes with an interesting hidden tweak, too, in the form of a move to better organize the list of installed apps in the Start menu – which would be a welcome change as we’ve already observed.

You might also like...

• Microsoft’s embarrassment over Recall fiasco gets worse as Windows 11 feature becomes the butt of Apple exec’s joke
• Don’t make these 5 big mistakes when using Windows 11
• You can run Windows 11 seamlessly on Apple silicon Macs

The owner of X broke with tradition at other social media companies to support the former president, as he drives political conversation on his site.

While Apple did unveil a redesigned, customizable Control Center as part of iOS 18, the technology giant didn’t formally announce a significant upgrade for the iPhone’s flashlight. Well, the latter has now landed in the iOS 18 Public Beta that's available to try now.

For some time, we’ve been able to adjust the brightness of the iPhone’s flashlight from zero to four, but with iOS 18, Apple’s giving us a bit more control within a fun user interface. Essentially, you can now more minutely control the brightness of the flash and the focus (aka the angle). 

I’ve been using it since the first developer beta of iOS 18 shipped shortly after Apple’s WWDC 2024 keynote. Folks willing to give the Public Beta of iOS 18 a go can do so now, as that’s also shipping as of July 14, 2024. 

Apple clearly had some fun making this feature

(Image credit: Future/Jacob Krol)

Now, when you turn on the flashlight on any modern iPhone – say an iPhone 15 Pro Max, like I’ve been using – you’ll see a window overlayed on top. Within that window, you’ll find an illuminated flashlight, and its glow is shown above. 

The magic happens when you drag your finger to the left or right. You can control the focus or angle, and top to bottom lets you adjust the brightness. So, whether you’re trying to illuminate a darker space or maybe provide light for a photoshoot, you have greater control over the luminance. 

If you’re keen to make further adjustments, you can long-press on that smaller flashlight icon on an iPhone with a Dynamic Island to bring up full control. It’s been quite responsive in my testing, and this functionality is also made possible by the True Tone Flash on the back of the iPhone. 

In terms of control, you can adjust both the focus and the brightness at once by sliding your finger from top to bottom and left to right – or vice versa, in any direction – to see it both render with the on-screen animation and the actual glow emitting from your iPhone. 

(Image credit: Future)

One could argue that Apple gamified the flashlight experience, as it is fun to use, actually helpful, and an excellent party trick. I’ve demoed it to a few of my friends, and they’ve been impressed. It’s also neat to see that Apple is delivering further control and enhancing the experience of a relatively mundane iPhone feature.

Apple’s iOS 18 will officially drop for all eligible devices this fall, but if you want to try out the new flashlight or more headlining features, you can opt-in and install the iOS 18 Public Beta now. Of course, as it’s a beta, it is a preview of the final release, and you can expect some bugs along the way and potentially a reduction in battery life.

Furthermore, if you’re looking to try out Apple Intelligence features, those are not yet included in the iOS 18 Public Beta or even the Developer Beta. Still, you get other anticipated features like the redesigned Photos app, a new reader in Safari, and text effects in iMessage, among others.

You Might Also Like

• I've covered Prime Day for six years - here's everything I'm buying at ...

US Congresswoman Jennifer Wexton (D-VA) faced the loss of her voice due to Parkinson’s Disease-like Progressive Supranuclear Palsy (PSP), relying on common robotic-voiced text-to-speech tools. Now, thanks to AI-powered voice cloning provided by ElevenLabs, she can give speeches in her original voice despite PSP’s impact on her vocal abilities. 

After Wexton used the robotic-sounding speech on the House floor, ElevenLab reached out to Wexton’s staff. With Wexton’s approval, her team provided ElevenLabs with over an hour of audio clips from before her diagnosis. In just a few days, ElevenLabs successfully created a digital version of Wexton’s voice, capturing the cadence, tone, and timbre that were once characteristic of her speech, according to the congresswoman and her team. The new voice model was first showcased as Wexton addressed the House Appropriations Committee, providing a strikingly natural sound compared to her previous computer-generated voice.

“Since I first began using an augmentative and alternative communication – or AAC – device for speeches on the House floor and in committee, I’ve received an outpouring of supportive messages from disability rights advocates about the importance of demonstrating that just because people like me may not be able to use our voices in the same way, doesn’t mean our words are any less ours or any less important to hear,” Wexton said in a statement. 

"You don’t understand how important of an issue accessibility is until it’s you who relies on a walker, or an AAC device, like I have learned to over the past year. I hope that this new step of adopting an AI voice model can also be a moment to start a conversation about new and creative ways we can continue to empower people facing the kinds of health and accessibility issues as I have, and to show that our abilities do not define who we are.”

Your ears aren’t deceiving you—AI has allowed me to make a new model of my voice like it was before my PSP.I hope this helps show creative ways we can empower people facing the kinds of health and accessibility challenges I have and demonstrate our abilities don’t define us. pic.twitter.com/ztXDjCgi63July 10, 2024

Accessible Speech

 Wexton pointed out in her speech that an AI-generated voice represents not just a personal victory but also highlights the transformative potential of AI in assisting individuals with speech impairments. By restoring a more natural-sounding voice, the technology enhances Wexton's ability to connect and communicate more authentically with her colleagues and the public.

ElevenLabs has quickly made a name for itself in the AI voice industry. The company's technology leverages advanced algorithms to not only replicate a person's voice but also to modulate tone and inflection, creating a more lifelike and natural sound. AI can even bring back the voices of those who have since passed away. The company recently showcased AI versions of the voices of James Dean, Burt Reynolds, Judy Garland, and Sir Laurence Olivier for its Reader App.

The successful implementation of AI-generated voices underscores a broader trend of integrating AI into assistive technologies. These advancements not only improve the quality of life for individuals with disabilities but also open up new possibilities for inclusion and participation in various spheres of life. By enabling more natural communication, AI technologies may help bridge the gap between ability and disability.

You might also like

• How you can get (AI versions of) Judy Garland or Burt Reynolds to read to you
• This app can add AI narration to any site or text – here's how to make it work
• AI Ghosts: Shaping the future of grief tech

Claims that President Biden and his allies ordered the attack on Donald J. Trump, or that Mr. Trump staged the attack, started quickly and spread fast across social media.

OpenAI is developing a new project to enhance its AI models' reasoning capabilities, called “Strawberry,” according to documents first discovered by Reuters. The project is a key element in the efforts by OpenAI to achieve more powerful AI models capable of operating on their own when it comes to performing research.

According to the internal documents Reuters looked at, Strawberry is aimed at building an AI that will not only answer questions but search around online and perform follow-up research on its own. This so-called “deep research” trick would be a major leap beyond current AI models that rely on existing data sets and respond in ways that are already programmed. 

There aren't details on the exact mechanisms of Strawberry, but apparently, it involves AI models using a specialized processing method after training on extensive datasets. This innovative approach could potentially set a new standard in AI development. An AI that can think ahead and perform research on its own to understand the world is much closer to a human than anything ChatGPT or other tools using AI models offer. It's a challenging goal that has eluded AI developers to date, despite numerous advancements in the field.

Reuters reported that Strawberry, which was then known as Q*, had made some breakthroughs. There were demonstrations where viewers witnessed AI could tackle science and math problems beyond the range of commercial models, and apparently, OpenAI had tested AI models that scored over 90% on a championship-level math problem data set.

Think Ahead

Should OpenAI achieve its goals, the reasoning capabilities could transform scientific research and everyday problem-solving. It could help plug holes in scientific knowledge by looking for gaps and even offering up hypotheses to fill them. This would vastly accelerate the pace of discovery in various domains.

If successful, Strawberry could mark a pivotal moment in AI research, bringing us closer to truly autonomous AI systems capable of conducting independent research and offering more sophisticated reasoning. Strawberry is, it seems, part and parcel of OpenAI’s long-term plans to demonstrate and enhance the potential of its AI models.

Even after GPT-3 and GPT-4 set new benchmarks for language processing and generation, there's a big leap to autonomous reasoning and deep research. But, it fits with other work on the road to artificial general intelligence (AGI), including the recent development of an internal scale for charting the progress of large language models.

You might also like...

• OpenAI has a new scale for measuring how smart their AI models are becoming – which is not as comforting as it should be
• If you think GPT-4o is something, wait until you see GPT-5 – a 'significant leap forward'
• ChatGPT just (accidentally) shared all of its secret rules – here's what we learned

The search giant’s negotiations to buy Wiz, a cybersecurity start-up, for $23 billion, come as the Biden administration has taken a hard line against consolidation in tech and other industries.

Windows 11 could be getting a significant alteration to the Start menu - and for a change, it’s a useful one, rather than some of the less welcome ad-related tweaks we’ve seen in recent times for the menu.

This was uncovered by leaker PhantomOfEarth on X, and it’s a new category view for the ‘All apps’ panel. However, this change is hidden in the new preview build 22635 in the Beta channel (released late last week), so it’s not officially underway just yet.

It looks like another Start menu > All apps view option is coming soon: Category view. Present in build 22635.3930, with some not yet functional categories.Also, you'll be able to switch between the existing alphabetical and new grid/category views using a dropdown menu. https://t.co/qFtcTm4BnB pic.twitter.com/ABtXZJ2IrlJuly 12, 2024

As you can see in the above post, this is another option for how to view your apps in the Start menu, as an alternative to the traditional alphabetical listing of applications – there’s also an incoming grid view (present only in testing), and now, the freshly discovered category choice.

This is essentially a grid of categorized apps, so you can have your apps sorted by category – such as entertainment, navigation and maps, news, productivity – and dive into those specific areas. As opposed to having to scroll down the long list of apps which is how Windows 11 organizes your installed software currently in the Start menu (a pretty clunky setup).

Analysis: Early work

This looks like very early work as when enabled (by using a Windows configuration tool with the mentioned preview build), the feature only shows colored blocks as the graphics for each category. We assume the four colored squares shown will host the icons of four apps from each category (maybe your most-used apps in every category, or that’s the first thought that occurs).

Of course, we don’t know how the interface will be realized, and this is just guesswork. We also need to firmly remember that as this is so early in the work towards this feature – it’s not properly implemented at all, and hidden in testing – it may not ever see the light of day in Windows 11.

After all, the grid view for apps in the Start menu is in testing as well, as mentioned, and so Microsoft might just decide a category view isn’t necessary – and that switching to a grid layout might be convenient enough for this part of the Windows 11 interface. Then again, more options don’t hurt, and the idea of organizing apps into categories might be an appealing way of rifling through installed software in the Start menu – we rather like this potential move.

Via Windows Latest

You might also like...

• Microsoft’s embarrassment over Recall fiasco gets worse as Windows 11 feature becomes the butt of Apple exec’s joke
• Don’t make these 5 big mistakes when using Windows 11
• You can run Windows 11 seamlessly on Apple silicon Macs