Technology

• Windows 11 has a new optional update
• It's a big one, delivering a whole load of new features
• These include an AI agent in Settings, a new quick-recovery option for PCs that won't boot, and added AI features for many Windows 11 apps

Windows 11 has a new update in preview, and it's a hefty download which packs in a lot of features, many of which are AI-related (and for Copilot+ PCs only, with their beefy NPUs).

This is patch KB5062660 for Windows 11 24H2, and it should be noted that this is an optional (preview) update, meaning that it's still in testing, so the various features may still have bugs (install it at your own risk, in other words).

Everything should be fully knocked into shape by the time the full update for August arrives, which is when all these features will be officially deployed (or should be – although some are on a 'controlled rollout', meaning they'll be drip-fed out).

So, with those caveats out of the way, what have we got to look forward to with this optional update (and next month's full upgrade)?

Here's my pick of the top features rolling out with KB5062660 (again, bear in mind that some are for Copilot+ laptops only).

1. AI agent in Settings

The headline functionality here is, sadly for most of us, for Copilot+ PCs only, and it's the addition of the first so-called Windows Agent. This is an AI agent specifically for the Settings app, and it lets you find and manipulate the options you need in a much more convenient way.

Normally, trying to find a setting involves using the search functionality, which can be rather hit-or-miss. With the AI agent, you're essentially getting an intelligent search where you can simply ask a question (in natural language) pertaining to what you're trying to do in Windows 11, and the agent will (hopefully) immediately surface the correct setting.

The AI doesn't just find the right setting for you, but can also make suggestions as to what changes you might want to make as well. Check out the video clip above of the agent in action to get a flavor of how it all works.

This is just rolling out to Arm-based Copilot+ PCs (with Snapdragon CPUs) to start with, but support for AMD and Intel chips is "coming soon" (it's also for the English language only initially).

Suffice it to say this is one of the more impressive uses of AI in Windows 11 I've seen so far, alongside more intelligent Windows 11 search from the desktop (for both Copilot+ laptops, and other PCs too). Yes, Microsoft appears to be progressing overall search capabilities nicely with AI, which is good to see.

(Image credit: Getty Images)2. Click to Do benefits

Click to Do is Microsoft's array of context-sensitive AI-powered options in Windows 11 (for Copilot+ PCs), and a few more shortcuts (for selected text or images) have been added here. That includes a choice to fire up Reading Coach, which is a free app (installed via the Microsoft Store) that aims to help you polish up your reading skills (pronunciation and more besides).

Secondly, Immersive Reader is now in Click to Do, which takes any text and presents it in a "distraction-free environment", allowing you to adjust text size, font, spacing, and a bunch of other parameters to make everything more easily readable. It also helps with reading skills (breaking down words into syllables, providing a picture dictionary, and more).

Draft with Copilot is also now in Word (for Microsoft 365 Copilot subscribers), allowing you to turn a sentence (or short summary) into a lengthy full draft penned by the AI.

(Image credit: xavier gallego morell / Shutterstock)3. Quick recovery from boot failure

Microsoft's Quick Machine Recovery (or QMR) is now arriving for all PCs (not just Copilot+ devices), which is the ability to recover from a problem that means your PC won't boot. This is the nightmare scenario we all dread, and the idea is that you'll be able to get help via the Windows Recovery Environment (which can be accessed if your system won't boot to the desktop).

QMR allows for diagnostic data to be sent to Microsoft and hopefully a patch can be sent back to cure the problem, or that's the idea. Another avenue of troubleshooting – one that's completely automated – is clearly a good idea, and I'm looking forward to seeing how useful this will be. (And I should clarify, I'm keen to see how competent this feature is – not to personally use it, with any luck. We can all hope that it's a screen we don't ever have to visit).

4. Image-related AI powers in Photos, Paint and Snipping Tool

A bunch of new AI-powered functionality for images is now inbound (for Copilot+ PCs), as previously seen in testing. That includes a Relight ability in the Photos app – allowing you to place virtual light sources to change the lighting in an image – and object select in the Paint app, which uses AI to select any given object (so you don’t have to do so manually).

A similar feature to the latter is coming to the Snipping Tool called 'perfect screenshot', which lets you roughly select an area of the screen you want to grab, and then automatically makes a precise crop of that element, taking the pain out of that process. (Again, check the above video to see how this works).

(Image credit: Microsoft)Other moves – including an important fix

There's quite a lot going on with this update, as mentioned, and another development is that Recall is now being deployed in Europe. Previously, this AI-supercharged search (that works using regularly taken screen grabs) wasn't out in the European Economic Area (EEA), but it is now, and it comes with a new ability. This is the option to export its screenshots (called snapshots) to "trusted third-party apps and websites" (wherever you want, basically - should you wish to share this data).

Furthermore, all Recall users are getting a reset button, which deletes all data relating to the feature and restores Recall to its original settings, should you wish to start afresh (or indeed abandon the ability and turn it off).

Also, the Black Screen of Death is now official, so wave goodbye to the blue version that's been with us so long, and say hello to a more streamlined effort whenever your PC is unfortunate enough to witness Windows 11 locking up. (I'm not sure about this change, as I've discussed elsewhere at length recently).

The Gamepad layout in the virtual keyboard for Windows 11 now benefits from "enhanced controller navigation" which includes word suggestions and better handling of menus, as well as the ability to use a gamepad to sign in from the Windows lock screen (via the PIN panel).

Finally, this preview update fixes a problem with the May 2025 update for Windows 11 which caused some PCs to suffer instability issues (crashes). Microsoft says this was a "rare" bug that didn't affect many, but it sounds like a truly nasty one, so having it resolved will doubtless be a relief.

You might also like...

• Microsoft promises to crack one of the biggest problems with Windows 11: slow performance
• Windows 11's handheld mode spotted in testing, and I'm seriously excited for Microsoft's big bet on small-screen gaming
• No, Windows 11 PCs aren't 'up to 2.3x faster' than Windows 10 devices, as Microsoft suggests – here's why that's an outlandish claim

• New ransomware proposals set to be implemented by UK Government
• These include a ban on ransom payments for public services
• Ransomware attacks cost the UK economy millions per year

A new set of cyber regulations are being introduced by the UK government in a move to protect British public services from the growing threat of ransomware.

Under the new guidelines, public organisations like the NHS, local councils, and schools will all be banned from paying ransom demands - a move which looks to dissuade criminals from targeting public institutions.

“Ransomware is estimated to cost the UK economy millions of pounds each year, with recent high-profile ransomware attacks highlighting the severe operational, financial, and even life-threatening risks,“ the Government’s statement argues.

Criminal business model

Amongst the high-profile attacks on public services in recent times is a ransomware attack on an NHS hospital, the disruptions from which contributed to a patient death - outlining just how damaging and harmful these attacks can be, not just financially but on a personal level too.

Ransomware-as-a-Service (RaaS) is a multi-million pound industry, and one which preys on vulnerable organisations without premium cybersecurity protections.

Public organisations in the UK with chronic budget restraints are particularly vulnerable, not just because they can’t afford first-rate security, but also because they provide critical services that have devastating impacts if disrupted.

Under the new set of proposals, businesses that are not covered by the ban are required to notify the government if they intend to pay the ransom - giving the government the opportunity to provide them with ‘advice and support, including notifying them if any such payment would risk breaking the law by sending money to sanctioned cyber criminal groups, many of whom are based in Russia.’

“Ransomware is a predatory crime that puts the public at risk, wrecks livelihoods and threatens the services we depend on,“ comments Security Minister Dan Jarvis. “By working in partnership with industry to advance these measures, we are sending a clear signal that the UK is united in the fight against ransomware.”

You might also like

• Take a look at our picks for the best antivirus around
• Check out our choice for identity theft protection tools
• Data of all 6.5 million Co-op members stolen - CEO says she is 'incredibly sorry'

If you’re going to watch The Fantastic Four: First Steps during its opening weekend, you’ll also be able to catch the very first trailer for Avatar 3 (better known as Avatar: Fire and Ash). As if I didn’t think my job could get any better, I was treated to an exclusive first look at what’s to come in the next installment of James Cameron’s smash-hit sci-fi series, in 3D, with a cocktail shaped like a delicious fiery explosion.

It was a similar fanfare for the release of The Way of Water, but I still couldn’t believe what I was seeing. Objectively, it seemed absurd for a trailer to have its own launch event, but as soon as the lights went down and I put my 3D glasses on, I completely understood why. Even in a matter of minutes, Cameron creates something so majestic, something so beautiful that stops time itself, that you can’t do anything other than sit back in amazement.

Obviously, I’m not going to tell you any specific details about the scenes in the Avatar 3 trailer – you’ll need to trust me when I say you need to see it to believe it. But I can tell you more about what to expect, and when you need to pay extra attention in order to work out what’s to come when the full film is released on December 19, 2025.

The Avatar 3 trailer is going to blow your mind, and the action is about to get a lot darker

Neytiri (Zoe Saldaña) in Avatar 3. (Image credit: 20th Century Studios)

Here’s the genius about the Avatar 3 trailer launch. I watched the footage through, in 3D, three times, and I still haven’t got the vaguest idea what the core storyline actually is. Essentially, this means the trailer is functioning as it should, not giving away everything we’re going to see in the full film, but simply alluding to it instead. But even if you can’t remember the ins and outs of Na'vi lore, it only takes a few seconds to work out that the next installment is going to be a lot darker.

As we already know, conflict on Pandora is about to escalate once again, with Sully (Sam Worthington), Neytiri (Zoe Saldaña) and the rest of the family encountering the aggressive Mangkwan Clan for the first time. They’re the “fire Na'vi,” and they’re viciously on the offensive from the moment they arrive in the trailer. We don’t know what their backstory is or why they have such an issue with outsiders, but as the trailer confirms, the meeting isn’t going to end in flowers and rainbows.

While the two Na'vi clans battle it out, Pandora itself has never looked more jaw-dropping. You can expect some ridiculously beautiful set pieces from Cameron in the trailer, elevating the 3D experience we got back in The Way of Water. It boggles the mind to think how they got there, but somehow, Cameron and his creative team has made something more visually astonishing that anything we’ve ever seen before.

What you need to watch out for

Sully (Sam Worthington) in Avatar 3. (Image credit: 20th Century Studios)

Even though I was as busy as everyone else getting lost in the wonder of Pandora, there was a particular scene that caught my attention, and potentially has the power to change everything we know about the Avatar franchise. Blink and you’ll miss it, because it’s on screen for less than a full second. Again, no outright spoilers here, but all you need to know is that the only familiar character involved in this fleeting moment is Sully.

From the first Avatar movie, we’ve known there’s a huge human element in Pandora’s world now, and you only need to look at Sully himself to understand how that’s adapted over time. However, it’s easy to forget that presence in Avatar 2, shifting the worldbuilding focus to different Na'vi tribes coming into contact. All I’ll say is that Avatar 3 has the power to bring the Na'vi vs. Na'vi vs. humans hate triangle back into play, all by introducing an entirely new perception of human popular culture into the mix (you’ll know it when you see it, I promise).

If 3D isn’t your thing, you’ll be able to catch the Avatar 3 trailer in full on YouTube from July 28. If you can’t wait that long… I suggest you get yourself down to your local cinema quick sticks.

• Stream all Avatar movies in the US on Disney+
• Stream all Avatar movies in the UK on Disney+
• Stream all Avatar movies in Australia on Disney+

You might also like

• I watched Avatar on the Apple Vision Pro and this is how you do 3D movies at home
• You can finally stream the original Avatar in glorious 4K HDR on Disney Plus
• 3D movies are great, actually

• Roku Channels adds 17 new channels
• Channels for wellness, reality TV, music, plus MGM's Stargate and American Gladiators
• Ambient TV channels combining visuals and low-key tunes

Roku has announced the addition of 17 new channels to its already huge selection, and that's terrible news for the productivity of home workers like me: there's an entire channel dedicated to American Gladiators, and another for Xena: Warrior Princess.

That's terrible news for my future paychecks but it's pretty good news for Roku subscribers, because the claim that there's something for everyone is actually true here: in addition to Gladiators and Xena, there are gripping documentaries, exciting music, and an entire channel dedicated to dogs.

What new channels is Roku adding?

Roku has already passed the milestone of offering more than 560 channels, and the 17 new additions are (thanks to Cord Cutters News for the channel numbers):

• 204 Swerve Sports: "the best in women’s football, basketball, soccer, volleyball, softball, gymnastics, boxing, rugby and more."
• 205 Team Liquid: esports, with competition and commentary from the popular esports team
• 207 PFL MMA: the Professional Fighters League deliver bone-crunching mixed martial arts
• 329 Xena: dedicated to Xena, Warrior Princess
• 576 Revolt Mixtape: hip-hop, R&B and much more
• 579 Cirque du Soleil: the legendary circus troupe deliver dazzling performances
• 586 Stingray Naturescape: "awe-inspiring landscapes, time lapses of serene night skies, aerial views of lush, tropical beaches, and the cozy glow of crackling fireplaces."
• 587 Zenlife by Stingray: wellness and meditation to help you relax
• 588 Stingray Stargaze: cosmic visuals, starry nights and ambient music
• 589 Stingray Cozy Cafe: lo-fi beats and, er, "inviting coffee shop interiors"
• 590 Stingray Cityscape: awesome aerial views set to down-tempo tunes
• 623 American Gladiators by MGM: MGM's take on Gladiators drew from both the US and UK versions, and was tons of fun
• 635 Paternity Court by MGM: edge-of-the-seat courtroom dramas drawn from real life
• 636 Life Below Zero: stories of surviving in some of the world's harshest environments
• 637 To Catch a Smuggler: real-life stories of law enforcement battling smugglers
• 638 Dog Whisperer by Cesar Millan: how to make bad dogs become perfect pooches
• 748 Stargate by MGM: superb sci-fi spanning the Stargate franchise

Sure, not all of these are going to be up everyone's alley – but I bet there's something there make you raise your eyebrows, even if it's just the idea of being able to watch Cirque du Soleil performers do something mind-blowing any time you feel like it.

The channels should be available right away on Roku TVs (including our favorite, the Roku Pro Series), and Roku devices such as the new Roku Streaming Stick Plus, which we recently gave a positive review to.

You might also like

• Don't worry if your Roku homepage suddenly changes – a smarter new layout is being tested for a lucky few
• Roku Streaming Stick review: the cheapest Roku streamer if you don’t need 4K
• Roku Streaming Stick Plus review: a budget 4K streamer that hides behind your TV

• Oracle's mystery $30 billion cloud deal is to support OpenAI and Project Stargate
• The project's first site, Stargate I, is already partly operational
• SoftBank, Microsoft and others continue to play a role

After weeks of speculation both with regards to Oracle's mystery cloud deal and potential delays to Project Stargate, OpenAI has confirmed the project is indeed still going ahead.

The deal, worth $30 billion per year, was revealed in a SEC filing by Oracle on June 30, but the company failed to mention who the big contract had been struck with.

We now know the contract involves 4.5 gigawatts of data center capacity, which equates to two Hoover Dams' worth of power or the equivalent of four million homes.

OpenAI confirms Stargate is still on

This brings the AI giant's total Oracle data center capacity up to five gigawatts, which is enough to power two million chips.

The huge deal with Oracle marks a major step forward in OpenAI's journey to 10 gigawatts, as outlined in its four-year, $500 billion Project Stargate strategy.

OpenAI confirmed its first site, Stargate I, will be located in Abilene, Texas, with Oracle and OpenAI jointly building it.

According to the company, parts of Stargate I are already operational, with early AI training and inference underway using Nvidia GB200 superchips.

The first site has already generated "thousands" of jobs, with OpenAI hoping to support 100,000 workers as part of the broader scheme.

Responding to reports that discussions had stalled between SoftBank and OpenAI amid disagreements, the ChatGPT-maker confirmed: "Our partnership with SoftBank is moving forward with strong momentum."

"Over the past six months, growing interest in Stargate has helped shape and expand our vision for this effort," OpenAI added. Despite broadening its list of providers and partners, the company added an interesting caveat: "Microsoft will continue to provide cloud services for OpenAI, including through Stargate."

Microsoft used to have exclusivity, but that all changed when OpenAI started to bring other cloud giants onboard.

You might also like

• These are the best cloud computing providers
• Check out out roundup of the best AI tools and best AI writers
• Oracle is reportedly giving the US Government some major software cloud discounts

Revealed at WWDC 25, macOS Tahoe 26 (which still feels strange to say) is promising big things. For one, there’s another visual refresh to match the Liquid Glass UI design of its OS siblings, transparent widgets and all.

Then there’s third-party Control Center actions support, a phone app, and, uh, Image Playground.

One feature above so many others stood out, though, and that’s the arrival of the Journal app on Apple’s oldest platform.

Day One of Forever

If you’d already assumed Journal was on macOS, then you’re not alone, as I saw a few bemused reactions to the announcement.

You see, Journal began life in iOS 17, and somehow never made it off of the iPhone. While it certainly had its strengths, creating a digital diary required using a fiddly phone keyboard and accessing the photos and videos stored locally on the device.

Ever tried to attach a file from an event to a Journal entry while using an iPhone? It’s not fun, and so my interest waned.

Journal is fine on iOS, but I just don't think phones are well-suited to the journaling process. (Image credit: Future / Apple)

Sure, I could fork out for Day One (and it’s a fantastic app), but I’m not at the point in my journaling journey (try saying that five times fast) where I can reliably use it for any more than a few days of the month, which would render any monthly subscription less than ideal.

Thankfully, macOS getting Journal means that it just got a lot easier to drag and drop to my heart’s content. Pictures and videos of a gig I’ve been to? Easy peasy. A PDF version of an event itinerary I want to look back on? Done, no more fiddly file management.

Then there’s that all-important physical keyboard, which means I’m much more likely to actually jot down some notes about how I’m feeling at the end of the day before I log off.

A different option

At the time of writing, I’ve been unsuccessfully trying to get my brain to jot things down in a Notion template I found online. And yet, when I look, it’s full of half-hearted workout records and random text snippets that are hardly worth keeping.

Sadly, having my journal tied into Notion - an app I use to run my freelance business - means every time I hop in to take some time for myself, I end up finding something I need to work on. After all, time is money, right?

Notion can be a powerful tool for productivity, but not so much for personal use. (Image credit: Future)

Jokes aside, it’s clear I need some sort of delineation between my journaling app of choice and Notion, which is essentially a work tool for me (and somewhere to catalog Magic: The Gathering cards, because a guy needs a hobby).

So could Journal on Mac be that?

But that’s not all!

Completing the home run for the Journal app on Apple platforms, it’s coming to iPadOS, too.

That means I can scribble my recollections in with the Apple Pencil, making it an even easier way to record my day-to-day adventures (that might be overselling sitting at a desk and writing).

It also gave me a great idea to ask my six-year-old son to write a few words every week so I could see his handwriting evolve and improve over time—something simply not possible on Apple’s other platforms.

Will you be giving Journal a go on macOS or iPadOS? Let us know in the comments below. I'm excited to restart my journaling process, and I'm hoping that Apple's signature Journal app will kick-start me into taking it more seriously this time around.

You may also like...

• Forget the Switch 2 – this dual-screen handheld gaming PC brings back the glory days of the Nintendo DS, but with all your Steam games
• Apple says it’s not merging iPadOS and macOS – but I reckon it might soon have to change its tune
• This restoration of the world’s oldest photograph shows the difference between AI slop and ChatGPT expertise

For a gaming mouse, the Cherry M64 is underwhelming.

Even though it has some compromises, I still love the Cherry XTRFY K5V2 keyboard.

• Proton has launched Lumo, a new AI assistant that's dedicated to your privacy
• There's a free version, and a paid-for Plus version with premium features
• You can use Lumo in a web browser or via apps on iOS and Android

Proton, the Swiss company behind apps including Proton Mail and Proton VPN, has launched its first AI product: Lumo.

Lumo is an open-source AI assistant that's designed with an emphasis on to privacy. You can use it for all the same things you would use ChatGPT for, like summarizing documents, helping with code or rewriting emails, but you'll be safe in the knowledge that everything you share with Lumo is completely confidential, and only readable on your device.

Proton guarantees that none of the information you give to Lumo will be used to train LLMs and can’t be shared with third parties.

I recently wrote about the dangers of unwittingly handing over sensitive information that's far more intimate than search or browsing history to AI companies, so Lumo's arrival is timely.

(Image credit: Proton)The dangers of Big Tech

Announcing Lumo, Andy Yen, Founder and CEO of Proton, said: “When we first launched Proton Mail a decade ago, our vision was to build a better alternative to Big Tech ecosystems for the benefit of all of society. Big Tech is using AI to supercharge the collection of sensitive user data to accelerate the world’s transition to surveillance capitalism.

"For this reason, we believe it is essential to provide an alternative that protects privacy and serves users as opposed to exploiting them. AI should not become the world’s most powerful surveillance tool, and our vision for Lumo is AI that puts people ahead of profits.”

So, if you’re looking for a European alternative to the big US and Chinese tech companies that are dominating the AI market right now, then Lumo could be a good idea.

(Image credit: Proton)

Lumo launches with key tools that we’d expect in an AI these days, including a web search, which is turned off by default for privacy but can be turned on. It has the ability to upload and analyze files, but unlike other AI assistants, no record is kept of any documents you upload.

Finally, Lumo has Proton Dive integration for adding files to chats. The files will stay end-to-end encrypted.

You can use Lumo for free, or upgrade to Lumo Plus for $12.99 a month (about £9 / AU$20). Lumo Plus gives you premium features including unlimited chats, extended chat history, unlimited chat favorites, and the ability to upload multiple and large files to Lumo.

You can sign up to Lumo now at lumo.proton.me, and there are apps for iOS and Android. Since it's only just launched we haven’t had a chance to compare Lumo to the more established chatbots like ChatGPT and Gemini yet, but you can be sure we’ll be putting it through its paces soon enough. Watch this space.

You might also like

• Spotify had to pull an AI-generated song that claimed to be from an artist who passed away 36 years ago
• This ChatGPT prompt will help you finally get started on the passion project you’ve put on hold for years
• Google should be worried – ChatGPT users now send 2.5 billion prompts a day, here’s how that compares

• A report suggests the foldable iPhone will have a 7.8-inch foldable screen and a 5.5-inch cover screen
• We've heard those sizes before, so they may well be accurate
• That though would make the foldable iPhone's screens significantly smaller than those of its main rival

There are still a lot more questions about the foldable iPhone than answers, and one of those questions is what size screens it will have. But we aren’t left wondering this due to a lack of answers – rather there have been many conflicting answers.

Now though, we’re starting to see some consensus, as a report from TrendForce (via MacRumors) claims that the foldable iPhone will have a 7.8-inch foldable display and a 5.5-inch cover screen – which are both dimensions we’ve heard before.

Ming-Chi Kuo (an analyst with a good track record for Apple information) said the same back in March, and we also heard almost identical sizes from tipster Digital Chat Station in February.

So while not all sources agree, there are now enough leaks pointing towards these sizes that they seem the most likely.

Smaller than the competition

The Samsung Galaxy Z Fold 7 has much bigger screens (Image credit: Lance Ulanoff / Future)

If these sizes are correct, then the foldable iPhone’s screen wouldn’t be especially large. The Samsung Galaxy Z Fold 7 for example has an 8.0-inch foldable screen and a 6.5-inch cover display. That means you essentially have a small tablet and a medium-sized smartphone all in one, whereas the foldable iPhone could end up with a foldable screen that’s still a fair bit smaller than a tablet’s, and a cover screen that’s quite compact.

We’d argue that might hold it back, since it probably wouldn’t fulfill either smartphone or tablet roles as well for most people, but then screen size is just one piece of the puzzle. If the remaining specs impress, the phone is suitably slim and light, and it’s not prohibitively expensive, then this could still be the device to make foldable phones mainstream.

We probably won’t find out for a while, as the foldable iPhone is unlikely to launch before late 2026 – and may arrive even later than that.

You might also like

• The foldable iPhone might be thicker than the Samsung Galaxy Z Fold 7, but that's okay
• The iPhone 17 Air might come in these 4 colors – and there's an interesting claim about its chipset too
• The iPhone 17 Pro Max could have a much bigger battery than any previous model

• Apple could add a sleep score in a future Apple Watch update
• It could measure your sleep stages, temperature, and more
• We don’t know when – or if – this feature will be added to watchOS

The best Apple Watches can track many things, including several different aspects of your nighttime slumber. But one thing they can’t do right now is provide you with a score that indicates the quality of your sleep. Yet according to a leaked graphic, that’s something that might soon be coming to Apple’s wearable.

That information was discovered by writer Steve Moser, who dredged up a graphic named “Watch Focus Score” from deep within the code of Apple’s Health app (via MacRumors). The combination of the image’s name and its contents might imply that Apple is working on a new sleep score feature for watchOS.

The picture depicts an Apple Watch with the number 84 in the center of its display. This number is surrounded by three bars that curve to form a circle. Interestingly, the bars are colored red, light blue and purple, and these tones correspond to the sleep stages shown in the Health app (there, red indicates time awake, light blue means REM sleep, and purple means deep sleep. The app also uses dark blue for core sleep, which could be what the graphic is showing).

The number and colored bars might hint at an overall score that takes into account the different sleep stages and how much of each you got at night. That would provide an extra level of data that you don’t currently find in watchOS.

More than just sleep stages?

(Image credit: Future / Britta O'Boyle)

But there are indications that other factors could be considered for this score. In Apple’s graphic, the Apple Watch is flanked on both sides by various icons, including a moon at stars, a “zzz,” a bed, and an alarm clock. Right now, Apple uses the bed icon for the sleep focus mode, while the alarm clock may signify when your alarm went off or when you got out of bed.

Moser also spotted a thermometer icon, which could be a hint that Apple will take more than just sleep stages into account when calculating a sleep score. It might incorporate wrist temperature as an indicator of your health, for example, and there may be other as-yet-unknown metrics that are also included as part of the overall score.

If this sleep score feature becomes a reality, Apple will be far from the first smartwatch maker to include it in their products: both Fitbit and Garmin have included sleep scores in their devices for years.

But Apple fans won’t mind that if they do indeed get this functionality in a future update – you never know, it might come to watchOS 26 later this year.

You might also like

• The Apple Watch is getting a sleep alarm upgrade it probably should have had 10 years ago
• The watchOS 26 public beta is imminent – here are 5 great free upgrades every Apple Watch will get, plus 4 you could miss out on
• Apple watchOS 26: Everything you need to know after WWDC

You can feel it in the silence, after the announcement, “We’re rolling out AI. It’s going to change everything.” No excitement. Just a quiet recalibration. More meetings. More tools. More disruption. Again.

For many organizations, AI isn’t landing as a breakthrough; it’s landing as a burden. Not because the technology doesn’t have potential, but because the way it’s being implemented is exhausting people. And exhausted people don’t drive transformation. This is what transformation fatigue looks like, and in the age of AI, it’s more common than ever.

AI’s problem isn’t the tech. It’s trust.

Across industries, teams are buckling under the weight of initiatives that arrive fast and land flat. With big promises, buzzwords and a new “strategic pivot” every quarter, under the surface, something deeper is breaking, having trust in the process.

Fatigue isn’t just exhaustion from doing too much, it’s frustration from doing too much that doesn’t matter. And AI, for all its promise, is becoming the latest culprit. When AI tools are introduced before teams are prepared, and when outcomes are measured in jargon, not value, enthusiasm evaporates.

Why product thinking cuts through the noise

This isn’t just a change problem, it’s a design problem. Today, too many organizations still treat transformation as a project. But AI doesn’t work that way, rather it evolves and iterates, it needs to be adopted in the flow of work, not bolted on.

This is where a product-led mindset makes the difference. In a product-centric operating model, change is continuous, and teams are cross-functional and close to the customer with value being delivered incrementally. And outcomes, not activities, guide decisions.

For IT management teams in particular, this shift is critical, they are often the first to feel the friction, implementing systems without full buy-in, training people on tools that weren’t designed with them in mind. These functions carry the weight of cultural change, yet are frequently excluded from strategic planning until rollout is already under way.

However, most organizations aren’t ready. A Harvard Business Review study found that 59 percent of product managers lack the skills to manage AI-driven products. To close the gap, 73 percent of companies are launching internal training, and those who do report a 28 percent increase in product success rates. It’s not the tech that makes AI work, it’s the capability around it.

What transformation fatigue actually looks like

The signs of fatigue aren’t always obvious, but they are almost always cultural.

One of the main causes of transformation fatigue is the long wait for value. AI initiatives often take too long to show impact, and belief in the cause drops off – teams disengage before results arrive. Then there’s the sense that new change looks suspiciously like the old change, leaders rebrand and employees begin to roll their eyes. In the end, it feels like version five of the same plan.

On top of this, methodologies start replacing thinking. Progress is measured in process, not outcomes. Buzzwords like “agile”, “transformation”, and “AI” lose meaning. And when capability gaps appear, the burden of change falls on people least equipped to carry it.

This is especially visible among frontline managers. They’re asked to adopt new systems, support new processes, and keep performance on track – all without enough context, training or time to adapt. The result isn’t just inefficiency, it’s disillusionment, which causes talent to walk out the door.

These are not just operational challenges, they are trust issues, and the longer they go unaddressed, the deeper the fatigue sets in.

So how do we fix it?

The importance of ownership

Companies should start with ownership, not just of tools, but of the transformation itself.

What this means is capability before rollout, organizing teams around delivering value, not around hierarchy, governing through experimentation, not perfection. It also means creating room for small failures, fast learning and constant adjustment.

Above all, it requires clarity. This means saying what’s changing, saying why it matters and making sure to say it again and again. Repetition isn’t the problem, confusion is.

This also means involving teams earlier in the process. Let them test, question and shape how change is applied in their context. Ownership doesn’t happen by decree. It happens through participation.

Transformation that actually transforms

Transformation fatigue isn’t inevitable. It’s a signal that the way we’re leading change isn’t working. The good news is that we don’t have to keep doing it this way.

Product-led thinking gives teams a different path forward, one that doesn’t rely on perfect plans, but instead builds momentum through visible progress. It builds capability, creating feedback loops and gets people involved.

It also builds trust. Not through slogans, but through small wins that actually matter. When teams see impact, they stay engaged and when leaders follow through, people follow back. In the end, when experiments are welcomed, better ideas emerge.

When you design change to work for people, not just around them, AI becomes a tool for focus, not friction. It becomes something worth investing in and believing in again.

That’s when transformation stops being exhausting and starts being real.

We've listed the best product management software.

This article was produced as part of TechRadarPro's Expert Insights channel where we feature the best and brightest minds in the technology industry today. The views expressed here are those of the author and are not necessarily those of TechRadarPro or Future plc. If you are interested in contributing find out more here: https://www.techradar.com/news/submit-your-story-to-techradar-pro

Are you seeking cheap internet options to help keep your budget down? Here are our top internet recommendations for older adults, seniors and retirees.

The AI hype felt relentless in 2023/24. While the initial frenzy has subsided somewhat, executives and professionals now grapple with the reality of deploying Artificial Intelligence (AI), specifically Generative AI (GenAI), within their organization.

LLMs (Large Language Models), the technology behind popular GenAI chatbots, are powerful, but there remains a significant disconnect between the perception of what they can do and their practical application for business writing.

Easy to use interfaces like ChatGPT make GenAI seem like it "can literally do anything".

This is a dangerous misconception. While incredibly useful for certain tasks, GenAI chatbots can be totally useless, and even harmful when not used appropriately.

Fundamental differences

The fundamental difference lies in how GenAI works compared to traditional software.

1. Traditional software is deterministic

It follows fixed logic and algorithms, producing the exact same, 100% accurate, and therefore repeatable result every time you give it the same input. Think of hitting CTRL+F in Word – you get a precise, repeatable count of a term.

2. Generative AI is non-deterministic

LLMs predict the next word based on probabilities from their training data. This means asking the same question twice will often give you different answers. They are designed to be variable.

Critical characteristics to understand

This core difference results in two critical characteristics businesses must understand:

1. Hallucinations: GenAI can confidently generate incorrect information or make things up. This isn't a bug; it's how the technology works. It's guessing based on patterns, not verifying facts. Copilot, for example, can wildly miscalculate readability scores or miss most instances of a search term.

2. Lack of Repeatability: You simply cannot guarantee the same output from the same prompt.

Here is the absolute critical takeaway: if your writing or document review task requires 100% accuracy or 100% repeatability, you must use deterministic software, not GenAI. Using GenAI for tasks demanding precision is a classic case of wielding a "GenAI hammer" and seeing every problem as a nail.

Flaws and errors in practise

Consider the disastrous consequences. I’ve used MS Copilot to search for every instance of "cybersecurity" in a contract for compliance purposes, only for the GenAI tool to miss 23 out of 27 occurrences. Trying to "shred" a document line-by-line into an Excel matrix for compliance, a task requiring perfect repeatability, is another inappropriate use case where GenAI will fail.

For businesses, especially in regulated sectors, using GenAI for tasks where factual accuracy is paramount is dangerous. Users may trust outputs due to brand credibility, not realizing the risks of inaccuracy.

Real-world failures like Air Canada's chatbot providing false information resulting in a lawsuit underscore the significant brand and trust damage inaccurate GenAI can cause.

So, where IS GenAI useful for business writing?

GenAI thrives for tasks where variability, creativity, or a "good enough" answer is acceptable or desired.

Appropriate use cases include:

• First Draft Creation: Generating initial versions of documents like management plans, executive summaries, or proposal sections based on context. This can save significant time.
• Creative Assistance: Rewriting content in a different tone or style.
• Summarization: Condensing lengthy documents.
• Simplification/Rephrasing: Making complex text more accessible or refining paragraphs.
• Research & Analysis: Using public data for competitive analysis or sales research where perfect accuracy on every detail isn't required for generating insights. Using NLP (another type of AI) for thematic analysis across communications to check message consistency.

Beyond simple chatbots, the real value often lies in specialized applications. These layer GenAI into workflows for specific jobs, intelligently combining GenAI for creative/drafting tasks with deterministic software for accuracy-critical functions like readability scoring or compliance checks.

They understand the "job to be done" and apply the right technology. NotebookLM, which generates audio summaries of documents, is a great example of a focused application.

Garbage In, Garbage Out: The Unsexy Truth of Knowledge Management

Generative AI, even when combined with techniques like Retrieval Augmented Generation (RAG) to access proprietary data, is not a magic wand that can overcome poor data quality. The old adage "garbage in, garbage out" is more relevant than ever. If your internal knowledge bases are a mess of outdated content, multiple revisions, and poorly tagged documents, the AI's output will reflect that chaos.

As the Harvard Business Review noted, "Companies need to address data integration and mastering before attempting to access data with generative AI". Good data hygiene – clear folder structures, naming conventions, and processes for maintaining content – is crucial but is fundamentally a human behavior problem, not just a tech one. Investing in proper knowledge management now will pay dividends when you roll out any GenAI solution.

Data Security: The Enterprise Achilles' Heel

Many popular AI chatbots rely on public cloud-based LLMs. For businesses, especially those in regulated industries like defense, finance, and healthcare, feeding proprietary or sensitive or PII (Personally Identifiable Information) data into these public models poses a significant security risk. CISOs (Chief Information Security Officers) are rightly wary, often blocking interactions with such models entirely.

The safer path for enterprises involves hosting LLMs in a private cloud or on-premise, fully locked down behind the firewall. The rise of powerful open-source models like Llama 4 or Mistral Nemo which can be deployed securely in-house, is a welcome trend. This shift is so significant that a Barclays CIO survey last year indicated 83% plan to repatriate some workloads from the public cloud, largely driven by AI considerations.

The Real Driver: People and Process

Most AI projects fail not due to the technology, but because of people, process, security, and data issues. Lack of buy-in, poor strategy, inadequate data, and insufficient change management and user education are common pitfalls.

Deploying AI chatbots without teaching users about:

• Hallucinations
• The need to verify outputs
• Effective prompting
• Crucially, what tasks not to use GenAI for

...will lead to frustration and project failure.

Start with the business problem you need to solve, then map the appropriate technology to that job. Don't just chase the "shiny new tech". Define your goals, measure success (both quantitative and qualitative), and involve end-users early.

When evaluating vendors, look beyond captivating demos. Ask pointed questions about accuracy, repeatability, data handling, security posture, and their understanding of your specific use cases and industry needs. Always try before you buy and vet vendors carefully. Be wary of vendors who overpromise or claim GenAI can do everything.

In summary, popular AI chatbots offer exciting capabilities, but they are not magic. They are powerful tools with significant limitations. Successful businesses will adopt a pragmatic, thoughtful approach: understanding GenAI's non-deterministic nature, applying it strategically to appropriate tasks (like creative drafting), leveraging hybrid applications, investing in data quality and security, and crucially, focusing on the people and processes required for effective adoption and change management.

This is the path to truly unlocking AI's value.

I tried 70+ best AI tools.

This article was produced as part of TechRadarPro's Expert Insights channel where we feature the best and brightest minds in the technology industry today. The views expressed here are those of the author and are not necessarily those of TechRadarPro or Future plc. If you are interested in contributing find out more here: https://www.techradar.com/news/submit-your-story-to-techradar-pro

• PayPal World will enable users to pay using their domestic wallets
• PayPal/Venmo, NPCI, Tenpin and Mercado Pago join forces
• The system works with open source APIs, so should be easily expandable

PayPal has launched a new platform in the hope of simplifying cross-border commerce by connecting major digital wallets and payment system to facilitate multi-currency transactions.

With PayPal World, users will be able to pay internationally using their domestic wallets and payment methods.

With over two billion users targeted globally, the new platform is expected to land in Fall 2025, but only select countries will be able to benefit from easier cross-border payments at launch.

PayPal World

From launch, PayPal World will work with Mercado Pago (Mexico), NPCI International Payments (India), PayPal (US), Tenpin Global (China) and Venmo (US).

"For much of the world’s population, international shopping and money transfers are not just difficult, at times they are impossible," the company noted in a press release.

The platform uses open source APIs to make it easy for more wallets to get onboard and boost interoperability in the future, but PayPal didn't mention any companies that could be joining its World platform beyond the initial launch partners.

"The challenge of moving money across borders is incredibly complex, and yet this platform will make it so simple for nearly two billion consumers and businesses," PayPal CEO Alex Chriss noted.

PayPal gave some examples of how global customers could use its new platform, including international visitors in China being able to scan their PayPal app with a merchant that accepts Weixin Pay, and UPI users in India being able to pay with their local wallet on an American ecommerce site.

NPCI International Payments CEO Ritesh Shukla welcomed the new platform, adding that it "aligns with [NPCI's] vision to make cross-border payments more seamless, secure, and inclusive."

"In addition to payments, Tenpay Global will deepen its collaboration with PayPal World in remittances," Tenpin Global CEO Wenhui Yang added.

You might also like

• This PayPal scam exploits new address feature to send out phishing scam emails
• These are the best payment gateways to support your shop
• We've listed all of the best money transfer apps

The hosts look to wrap up a series win at Old Trafford Cricket Ground.

Commentary: This series walked so The Boys could fly.

• Running an old OS like Windows 10 could soon pose cybersecurity issues
• Upgrading could reveal software compatibility issues, report notes
• The clock is now ticking to avoid a rushed Windows 11 migration

New research has uncovered some of the finer details around why many businesses are still being cautious with their approach to Windows 11 migration, with security threats and financial impacts proving to be major hurdles.

The report from Panasonic found nearly two-thirds (62%) of devices need replacing or upgrading for Windows 11 compatibility, highlighting the scale of the problem – a figure that rises to 76% among larger organizations with 5,000+ employees.

However, despite migration-related concerns, the study claims many organizations still recognize the benefits of upgrading from Windows 10 and older operating systems.

Businesses still have some concerns about upgrading Windows

Panasonic found 94% fear increased ransomware and malware risks if they don't upgrade, with 93% also concerned about data breaches. But two in three noted overall higher costs associated with migrating to Windows 11, with 55% stating that it could add to cybersecurity expenses.

Nearly half also noted software compatibility issues (47%) and productivity loss during downtime (45%), and for many (25%), hardware upgrades come with software upgrades, compounding the financial impact of OS upgrades.

However, with Microsoft estimating that ESU could cost around £320,000 over three years for 1,000 devices, the need to upgrade is clear.

Around a third each acknowledge that upgrading will give them better performance and processing power (36%), a more future proof ecosystem (36%) and access to AI features like Microsoft Copilot (34%).

Panasonic TOUGHBOOK Europe Head of Go-to-Market Chris Turner commented: "The window is closing for organisations to make a well-planned, measured and cost-effective transition to Windows 11 and start unlocking its benefits."

"Organisations that are still to undertake Windows 11 migration need support to ensure their deployment is not rushed and risky," Turner added.

You might also like

• Windows 10 users who don’t want to upgrade to Windows 11 get new lifeline from Microsoft
• We've listed the best business laptops and best mobile workstations for the best portable upgrades
• Check out our roundup of the best productivity software

California-based Lucid Motors completes its transition to the NACS charging standard via a $220 adapter.

• Leaked renders have shown the Pixel 10 Pro and Pixel 10 Pro XL in four shades
• These include Obsidian, Porcelain, Moonstone, and Jade
• Two of these are far more interesting than the other two

It’s looking likely that you’ll be able to buy the Google Pixel 10 Pro and the Pixel 10 Pro XL in a choice of Obsidian (black), Porcelain (white), Moonstone (slate blue-gray), and Jade (a soft pistachio green with gold accents), as not only have some of these Pixel 10 colors been mentioned before, but now all four have been shown off in leaked renders.

Android Headlines has shared what it claims are official renders of the Pixel 10 Pro and the Pixel 10 Pro XL in these four shades, and while we’d take this leak with a pinch of salt, these certainly look to be high-quality images, so they may well be official.

If these renders are accurate, then the Pro models will be available in two fairly plain, ordinary shades, in the case of Obsidian and Porcelain, since they’re basically just black and white. But the other two options are a bit more interesting.

Image 1 of 2

Leaked renders of the Pixel 10 Pro in four colors (Image credit: Android Headlines)Image 2 of 2

Leaked renders of the Pixel 10 Pro XL in four colors (Image credit: Android Headlines)A bit more color

There’s Moonstone, which we’ve actually seen the Pixel 10 Pro in already via an official teaser. This is rather understated, but the hint of blue in it makes this more interesting than a pure gray option.

The highlight, though, is arguably Jade – it’s a soft, delicate shade that still somewhat fits with the rest of the color options, but is a bit brighter and more unusual. Really, we’d like to see more of this sort of thing, rather than top-end phones defaulting to plain shades, but at least there’s one option here for those who want a splash of color.

We’ll find out how accurate this color leak is soon, as Google is set to unveil the Pixel 10 series on August 20. We’re expecting to see the Pixel 10 itself along with the Pixel 10 Pro, the Pixel 10 Pro XL, and the Pixel 10 Pro Fold, so there should be a lot to look forward to.

You might also like

• Google just teased the Pixel 10 Pro – and a fresh new Pixel color
• The Google Pixel 10 series just appeared on multiple official databases, suggesting an imminent arrival
• The Google Pixel 10 Pro Fold could beat the Samsung Galaxy Z Fold 7 in two significant ways