Posted on

Microsoft Recall still captures SSNs, passwords, and more

Microsoft had to recall the Recall AI feature on Copilot+ Windows PCs earlier this year. In theory, Recall was supposed to be a great AI trick to quickly resurface information from your past activities on a computer. In practice, it turned out to be a privacy and security nightmare.

Researchers proved that Recall data could be easily accessed by a malicious actor who gains access to a computer. Also, the saved screenshots could contain sensitive information, and Recall was enabled by default on those Windows builds.

Microsoft heard the complaints and pulled the feature to rework its security. As testers have found, Recall now encrypts the screenshots, making it impossible to extract information from the database. However, Recall will still take screenshots of sensitive data, including social security numbers, logins (including passwords), and credit card data.

Even with protections in place, Recall mostly failed to recognize that sensitive data was entered on the screen in Tom’s Hardware’s tests.

Tech. Entertainment. Science. Your inbox.

Sign up for the most interesting tech & entertainment news out there.

By signing up, I agree to the Terms of Use and have reviewed the Privacy Notice.

The feature has a “filter sensitive information” setting that everyone should enable right if you use Recall. The good news is that Recall is an opt-in service now, meaning it won’t be automatically enabled once you get the latest Windows update.

The test showed various scenarios that many computer users will reconize where Recall shouldn’t take screenshots of the content shown on the display. But Recall captured the information anyway.

For example, Recall recorded a Windows Notepad window where the user entered a credit card number and a random username and password combination, even though the user typed “Capital One Visa” next to the number to trigger the protections. Recall also captured fake social security numbers, names, and date of birth details entered into a PDF loan application. The feature had no idea those were all fake; it just screenshotted the information. It also captured the page when a genuine credit card number was introduced.

The tester then created a web form with fields that said, “Enter your credit card number below.” The form asked for the credit card type, number, CVC, and expiration date. Again, Recall failed to trigger the protections and recorded images containing the data.

The only time Recall worked correctly was when the user entered payment information into the forms of two online stores.

That’s certainly not good enough and a big reason to worry. I’m a longtime Mac user, so I won’t deal with Recall anytime soon. Or if I do, I’ll only do it for testing purposes. But I can’t see why one should leave it enabled.

The report notes that Recall’s databases are indeed encrypted. Also, you need to authenticate with Microsoft Hello (fingerprint, face, or PIN) to access your screenshots. But a malicious actor with remote access to your computer could get into the history the AI uses once they obtain that PIN. You might think it won’t happen to you, but I wouldn’t take that risk, no matter how amazing this Copilot feature might be.

Because yes, I won’t lie, having the AI remember what websites you visited can be an amazing tool if done right. The feature would have to have strong security and privacy features, and Microsoft is not getting it right.

When asked about Recall’s feature that’s supposed to identify sensitive data, the company offered Tom’s Hardware a quote from its Recall blog post that tackles privacy matters:

We’ve updated Recall to detect sensitive information like credit card details, passwords, and personal identification numbers. When detected, Recall won’t save or store those snapshots. We’ll continue to improve this functionality, and if you find sensitive information that should be filtered out for your context, language, or geography, please let us know through Feedback Hub. We’ve also provided an option in Settings that we encourage you to enable that will anonymously share the apps and sites you prefer to be excluded from Recall to help us improve the product. And you can also choose to exclude specific apps and websites through the Recall settings page which we talk about below.

That means Recall will get better over time, but you’ll have to wait. Until that happens, it might capture sensitive data while you use your computer, as these tests have shown. You’ll have to decide for yourself whether the risk is worth it.

You can try Recall by installing the newest Windows Insider Build. Read more about it in Microsoft’s blog at this link.

Source

Posted on

Government review of denied datacentre builds sees Iver project get green light

A government review of a local council’s decision to block a US-based company from building a hyperscale datacentre in Iver, Buckinghamshire, has concluded the project should proceed.

Buckinghamshire Council refused permission in November 2022 for US investment company Affinius Capital to proceed with its plans to redevelop an industrial estate in Court Lane, Iver, Buckinghamshire and build a 65,000m2 datacentre on the site instead.

The reason given by the council for the refusal is that the project would be an inappropriate use of Green Belt Land, which are protected pieces of land that are intended to prevent the onset of urban sprawl.

Shortly after coming to power in July 2024, the Labour government pledged to review the council’s decision to block the project in support of its strategy to stimulate the UK’s economic growth by accelerating the delivery of large-scale infrastructure projects.

The developer had raised an appeal against the council’s decision, and a month before the government’s intervention a public local inquiry was held over four days in June 2024.

Following a review of the council’s decision and the local inquiry, the government has now granted Affinius Capital permission to proceed with the project, with a letter dated 6 December 2024, outlining the reasons why.

The letter states that the decision to overturn Buckinghamshire Council’s decision to block the build was made by the minister of state for Housing and Planning Matthew Pennycook, on behalf of the secretary of state Angela Rayner.

“Weighing in favour of the proposal are the need for new datacentres, reduction in HGV movements, heritage benefits, reuse of previously developed land, and investment and job creation, which each carry significant weight,” the letter stated.

“Weighing against the proposal are harm to Green Belt, which carries substantial weight; harm to [a nearby] listed building, which carries great weight; and landscape harm and visual harm, which carries moderate weight.”

The letter also goes on to state that, in Rayner’s view, there are “very special circumstances to justify this development in the Green Belt”, adding: “The secretary of state therefore concludes that the appeal should be allowed and planning permission granted.”

The letter also states that the secretary of state’s decision on this matter can be challenged in the High Court, provided an application to do so is received within six weeks from the date of the letter.

Computer Weekly contacted Affinius Capital for comment on this story, but no response was received by the time of publication.

The Affinius Capital project was one of two datacentre developments the government placed under review in July 2024.  

The other is being overseen by Oxford-based developer Greystoke Land, after its bid to build a £1bn datacentre in Abbots Langley, Hertfordshire, was denied in January 2024. That decision is being appealed.

At the time of writing, Computer Weekly understands a decision at government level on whether that build will go ahead remains pending.

Source

Posted on

Uh oh… Zotac just leaked Nvidia’s next-gen launch line-up, including RTX 5090 GPU with 32GB of VRAM

  • Nvidia RTX 5090 has 32GB based on leaked info from Zotac’s website
  • Next-gen launch GPUs are supposedly the RTX 5090, 5080, 5070 Ti and 5070
  • There is, however, no sign of the RTX 5060 in this spillage

Zotac just leaked details of what might be the initial line-up of next-gen desktop graphics cards from Nvidia.

VideoCardz did the sleuthing here, turning up details Zotac accidentally aired on its own website, showing us the Blackwell GPUs that the graphics card maker will initially debut (in theory, anyway). Furthermore, Zotac also dropped a tasty nugget of info on the VRAM configuration for what’s surely the next-gen flagship.

The models listed by Zotac – and all the spilled details have now been removed, we should clarify – were as follows:

  • Nvidia GeForce RTX 5090
  • Nvidia GeForce RTX 5090D
  • Nvidia GeForce RTX 5080
  • Nvidia GeForce RTX 5070 Ti
  • Nvidia GeForce RTX 5070

The Nvidia RTX 5090D is the variant of the flagship for China, following in the footsteps of the RTX 4090D, as you’re likely aware.

As for the VRAM info, Zotac has filters for its GPUs to allow sorting by memory type and capacity, and mistakenly put a GDDR7 option in the former, as well as an allocation of 32GB in the latter.

This shows us that RTX 5000 graphics cards will carry GDDR7 VRAM as (heavily) rumored – all models will use this cutting-edge memory, supposedly – and that there’ll be a 32GB allocation of video RAM in the line-up, as there isn’t with the current-gen (which tops out at 24GB).

The GPU paired with 32GB must, of course, be the RTX 5090, and this is what’s already been rumored for the next-gen flagship.

{ window.reliablePageLoad.then(() => { var componentContainer = document.querySelector(“#slice-container-newsletterForm-articleInbodyContent-cjjioXmoF6tM27hYVun9X3”); if (componentContainer) { var data = {“layout”:”inbodyContent”,”header”:”Get daily insight, inspiration and deals in your inbox”,”tagline”:”Sign up for breaking news, reviews, opinion, top tech deals, and more.”,”formFooterText”:”By submitting your information you agree to the Terms & Conditions and Privacy Policy and are aged 16 or over.”,”successMessage”:{“body”:”Thank you for signing up. You will receive a confirmation email shortly.”},”failureMessage”:”There was a problem. Please refresh the page and try again.”,”method”:”POST”,”inputs”:[{“type”:”hidden”,”name”:”NAME”},{“type”:”email”,”name”:”MAIL”,”placeholder”:”Your Email Address”,”required”:true},{“type”:”hidden”,”name”:”NEWSLETTER_CODE”,”value”:”XTR-D”},{“type”:”hidden”,”name”:”LANG”,”value”:”EN”},{“type”:”hidden”,”name”:”SOURCE”,”value”:”60″},{“type”:”hidden”,”name”:”COUNTRY”},{“type”:”checkbox”,”name”:”CONTACT_OTHER_BRANDS”,”label”:{“text”:”Contact me with news and offers from other Future brands”}},{“type”:”checkbox”,”name”:”CONTACT_PARTNERS”,”label”:{“text”:”Receive email from us on behalf of our trusted partners or sponsors”}},{“type”:”submit”,”value”:”Sign me up”,”required”:true}],”endpoint”:”https://newsletter-subscribe.futureplc.com/v2/submission/submit”,”analytics”:[{“analyticsType”:”widgetViewed”}],”ariaLabels”:{}}; var triggerHydrate = function() { window.sliceComponents.newsletterForm.hydrate(data, componentContainer); } if (window.lazyObserveElement) { window.lazyObserveElement(componentContainer, triggerHydrate); } else { triggerHydrate(); } } }).catch(err => console.error(‘%c FTE ‘,’background: #9306F9; color: #ffffff’,’Hydration Script has failed for newsletterForm-articleInbodyContent-cjjioXmoF6tM27hYVun9X3 Slice’, err)); }).catch(err => console.error(‘%c FTE ‘,’background: #9306F9; color: #ffffff’,’Externals script failed to load’, err)); ]]>

Sign up for breaking news, reviews, opinion, top tech deals, and more.

An Nvidia GeForce RTX 4060 on a table with its retail packaging

(Image credit: Future / John Loeffler)

Analysis: What about the RTX 5060, though?

With this kind of work going on with manufacturer websites, in the background – well, it should have been on the quiet, in the background, but was accidentally sent live by a Zotac employee, clearly – shows we are about to get new RTX 5000 GPUs at CES 2025. Although Nvidia has all but said that, anyway, at this point.

The really interesting bit here is the underlining of the RTX 5090 being a mighty GPU sporting 32GB of video RAM, and the range of models available initially, which are as expected, pretty much. Well, the RTX 5090 and 5080 are, anyway, the rumor mill just isn’t quite sure if we’ll also get the RTX 5070 or the 5070 Ti – and maybe this is a suggestion that Nvidia will push out both. Alternatively, perhaps one of those RTX 5070 variants may come slightly later.

Notably, there’s no mention of the RTX 5060, which has recently floated up on the rumor winds as a possible GPU launch for later in the first quarter of 2025. Zotac may not be prepping that because it’s a couple of months down the line from these initial launches – or perhaps this is a hint that this lower-tier Blackwell graphics card won’t turn up until later in 2025.

You might also like

Source

Posted on

Google smart glasses with Gemini AI hands-on: Google Glass done right

When it demoed Project Astra at I/O 2024 in May, Google teased smart glasses for the first time. The company did the same thing during Wednesday’s big Gemini 2.0 announcement, where the wearable was part of a longer Project Astra demo. Google also suggested Gemini AI smart glasses might be coming soon.

At the time, I thought Google was simply showing off a prototype of a pair of Samsung XR glasses that would be unveiled next month during the Galaxy S25 Unpacked press event. Little did I know that Samsung and Google had bigger things in mind.

A day later, the two companies unveiled Project Moohan, a Vision Pro spatial computer rival. Google also announced the Android XR platform that will power it. Both Samsung and Google mentioned smart glasses in their announcements. But Google’s was more impressive, as the company showed off AR features for smart glasses powered by Gemini AI.

Google didn’t offer a name for the smart glasses or a release date. But the company did give a select few people a hands-on experience with Project Moohan and the unnamed smart glasses. It turns out the Gemini AI wearable is quite interesting, seemingly delivering the Google Glass experience that Google failed to offer more than a decade ago.

Tech. Entertainment. Science. Your inbox.

Sign up for the most interesting tech & entertainment news out there.

By signing up, I agree to the Terms of Use and have reviewed the Privacy Notice.

We all remember the Google Glass project, which was, in retrospect, ahead of its time. That smart glasses concept sparked privacy worries at a time when Google wasn’t exactly known for great privacy. Also, there was no generative AI at the time to truly make Google Glass useful.

Fast-forward to late 2024, and Google seems confident enough to demo a product that could become a must-have accessory for people who want AI assistance all the time. Wired tested Google’s Gemini AI glasses and Project Moohan, finding the former the more compelling product.

The glasses seem to take inspiration from the North Focals, smart glasses from a company that Google purchased a few years ago. But they’re slimmer and more comfortable than the Focals.

Even so, smart glasses have thick arms and thicker rims around the eyes, which is what you’d expect from glasses that incorporate AR abilities. They feature clear or sunglasses lenses and will support prescription lenses like Moohan.

North Focals glasses.North Focals glasses. Image source: North

When it comes to AR capabilities, the glasses come in three versions. The no-AR model is presumably the cheapest, as it lacks a display. Then there’s a pair that projects images on one of the lenses, the monocular display.

The best experience, and probably the most expensive, comes from the binocular display version that will center AR images, as seen in the photos above and below.

The processing is passed on to a nearby smartphone, presumably a Pixel phone that connects wirelessly to the glasses. That’s where Gemini resides, and the AI is ready to help with a tap on the arm of the glasses. A tap on the side also brings up the display in those models that support AR.

Google Pixel 9 Pro Fold GeminiGoogle Pixel 9 Pro Fold: Gemini support. Image source: Christian de Looper for BGR

The camera is also built into the frame, and an LED turns on when it is active. Built-in microphones pick up your commands for Gemini, and a speaker in the frames lets you hear the AI talk back.

The glasses are meant to last a full day on a charge, though battery life will probably depend on how much you use them. The hands-on experience with the Gemini AI smart glasses doesn’t offer actual battery characteristics, and it’s too early for that.

The report explains the various scenarios where AI smart glasses will be useful. For example, the glasses support Google Maps navigation, a feature Google showed in the Android XR platform announcement (image below).

Google Maps AR navigation on smart glasses.Google Maps AR navigation on smart glasses. Image source: Google

You also get AI summaries of notifications displayed in front of your eyes. The same goes for real-time translation of text. Impressively, the glasses can translate spoken foreign languages in real-time. Gemini will also caption conversations, a great feature for people with hearing issues. And Gemini can answer in multiple languages, a feature ChatGPT’s Advanced Voice Mode also supports.

The AR capabilities do not stop there. The glasses will show you previews of photos you take, and you’ll be able to play video when needed. The display experience isn’t the greatest, which is understandable. But it still sounds like the hands-on demo was a success.

Gemini powers all the smart AI features, like real-time translation and captioning. The AI can also summarize content seen while on the go, like the page of a book. Gemini has a short-term memory, too, so it can recall some of the things you’ve just done and seen a few minutes ago.

Interestingly, Gemini can also identify products and offer instructions on how to use them. In this hands-on, the Wired reporter asked how to use a Nespresso machine.

As impressive as the demo might sound, Google wasn’t ready to provide a release date and pricing information. The Gemini AI smart glasses it unveiled are clearly superior to the Meta Ray-Ban smart glasses, which only offer AI support, with the AR component missing. I’d expect Google’s glasses to be more expensive, assuming Google wants to sell them.

Since Project Astra is still in the early days, I’d expect Google to launch the glasses once the Gemini assistant abilities it’s working on for Project Astra are ready to launch. No point in having smart glasses in store if the Gemini software isn’t ready.

Meanwhile, you should check out Wired’s full hands-on experience for more details.

Source

Posted on

Post Office creates CTO role to support ‘extensive and complex’ plans

The Post Office has created a new role to bolster its tech leadership team amid an ongoing project to replace the Horizon system supplied by Fujitsu.

This comes at a time when the organisation faces tough decisions over the move away from the controversial IT system used in all Post Office branches.

Subpostmasters were blamed for unexplained shortfalls caused by errors in the Horizon software, which was introduced in 1999. Hundreds were wrongly convicted of crimes based on flawed evidence from the system, in what is known as the Post Office Horizon scandal.

Paul Anastassi joins the Post Office from gaming and entertainment company Rank Group as interim chief technology officer (CTO). Acting CEO Neil Brocklehurst told staff in an internal message: “…we are keen to transition to our new Executive Operating Model as quickly as possible in order to deliver on the benefits of the Strategic Review for Postmasters and enable broader change as necessary.”

He added: “With this in mind, we have taken the decision to appoint to the CTO role on an interim basis whilst we complete the recruitment process for our permanent CTO, which is now underway internally and externally.”

Andy Nice, chief transformation officer at the Post Office, told Computer Weekly: “The interim CTO role was introduced to allow us to increase our technology leadership capability given our extensive and complex plans for this area, as well as enabling me to focus on the delivery of our strategic transformation plan for the business.”

He said Anastassi has “excellent, relevant experience in creating and executing technology strategies to enable business improvement and growth”.

The Post Office – under its new leadership, including recently appointed Nice – is in the midst of a huge project to replace the controversial Horizon system from Fujitsu, which is at the centre of the Post Office scandal.

Nice and his team were quick to act on arrival at the Post Office earlier this year, pausing the work being done on its planned Horizon replacement, the New Branch IT (NBIT) project.

The NBIT project to build an in-house software platform to replace Horizon was running late and hugely over budget. Costs had increased by £1bn and, as revealed by Computer Weekly in May, a government report described the project as “unachievable”.

There is still an ongoing debate at the Post Office about the way forward for the project, with claims the Post Office is set to buy the Horizon system from Fujitsu and combine it with in-house developed and commercially available software.

The Post Office scandal was first exposed by Computer Weekly in 2009, revealing the stories of seven subpostmasters and the problems they suffered due to the accounting software (see timeline of Computer Weekly articles about the scandal below).

Source

Posted on

AWS on using GenAI to speed up legacy VMware and Microsoft datacentre migrations

Amazon Web Services (AWS) has set out how its investments in artificial intelligence (AI) chips and software are saving customers money and helping them migrate their legacy Windows and VMware workloads off-premise much quicker.

AWS CEO Matt Garman used the opening keynote at the public cloud giant’s Re:Invent customer and partner conference in Las Vegas, which is the first he has delivered since taking over the company reins in June 2024, to talk up the potential for generative AI (GenAI) to digitally transform the way that businesses operate. He also talked at length about the work that goes into ensuring the AWS cloud infrastructure is equipped to cope with the growing demand from its customers for the compute power they need to run AI and GenAI workloads.

As previously reported by Computer Weekly, the demand for GenAI workloads from its customers was recently cited as the reason for a “significant re-acceleration” in AWS’s annual growth rate, with the company reporting a 19.1% year-on-year uptick in revenue during its third-quarter results.  

Garman touched on Amazon’s 14-year-long collaboration with Nvidia, which he said has enabled it to roll out a succession of increasingly more powerful graphics processing unit (GPU) instances based on the latter’s technology so it can keep pace with its customers’ AI demands.

The company has also doubled down on the creation of its own AI silicon – namely its family of Tranium chips – to support a wider range of instances that are designed to improve the cost performance of running compute-intensive workloads. To this point, Garman used the keynote to announce that the second generation of Tranium instances had now become generally available, claiming the latest iteration can deliver “30-40%” better price performance than “current GPU-powered instances”.

This is based on feedback from early adopters of the technology, with Garman naming Adobe as among the customers who have seen some “promising” early wins with the technology.

Another is AI-focused software engineering startup Poolside, who has reportedly committed to training all future versions of their large frontier model on Tranium 2. The company is also anticipating the move will generate savings in the region of 40%. “Databricks is one of the largest data and AI companies in the world,” he said. “[It] plans to use Trainium 2 to deliver better results and [to] lower the total cost of ownership for our joint customers by up to 30%.” 

Opening up about Amazon’s use of GenAI

The conversation later moved on to how GenAI is also changing the way that AWS operates, with particular focus on how its own offerings are helping to speed up the time it takes to refactor legacy, on-premise workloads and ready them for migration to the public cloud.

Central to this bit of the discussion was Amazon Q, which is the company’s generative AI chatbot assistant that is designed for in-house use by software developers, business analysts and contact centre employees to make the work they do more efficient.

The migration of customer workloads out of private datacentres and into the public cloud is a process that fuelled the company’s growth for a decade or more after its inception in 2006.

However, despite the company previously acknowledging that a large proportion of enterprise workloads remain on-premise, it was an area that was markedly less talked about during the keynote, until Garman flagged how Amazon Q can assist with this task.

“Our goal at AWS is to help every builder be able to innovate, [and] we want to free you from the undifferentiated heavy lifting to really focus on those creative things that make your building unique … [and] generative AI is a huge accelerator of this capability,” he said.

As an example, he talked about how Amazon Q Developer, an iteration of the chatbot specifically designed to help developers speed up their CodeDeploy processes, is helping customers deploy faster, more secure and better-quality software updates.

Garman then went onto announce several new features that were being added to Amazon Q Developer that will generate unit tests, documentation and code reviews on behalf of developers, so they can spend more time each day writing code than dealing with the admin associated with it.

Addressing the legacy

The software is also reducing the amount of time they have to spend managing legacy applications, it is claimed.

“One of [the software’s] most powerful capabilities we already have is [its ability to] automate Java version upgrades,” said Garman. “What it can do is transform a Java application from an old version of Java to a new version in a fraction of the time it would take to do manually. This is work that no developer loves to do, but is critically important.”

According to Garman, integrating this capability into Amazon’s own internal systems saw it “migrate literally tens of thousands of production applications” to Java 17 in a “small fraction of the time” it would typically take. “The estimate from our teams is this saved us 4,500 developer years … [and] this is a mind-blowing amount of time saved, and because we’re now running on modern Java, we can use less hardware, too. So, we saved $260m a year through this process.”

Java upgrades are one thing, but – in Garman’s opinion – a migration that a lot of enterprises want assistance with is moving from Windows to Linux. And this is something AWS can assist with now through the preview release of a new version of Amazon Q Developer.

“Customers love an easy button to get off of Windows,” he said. “They’re tired of constant security issues, the constant packing or patching, all the scalability challenges that they have to deal with, and they definitely hate the onerous licensing costs.

“But we do recognise today that this is hard. Actually, modernising away from Windows is not easy, [but] with Q Developer, modernising windows just got a lot easier … [as it allows you] to transform .Net applications that are running on Windows to Linux in a fraction of the time.”

Signature IT

As an example, Garman flagged digital transactions, signing software company Signature IT, and the work it has done to modernise its legacy .Net applications and migrate them from Windows to Linux. “It was a project they estimated was going to take six to eight months, [and] they actually completed it in just a few days,” he said. “That is a game-changing amount of time.”

But it’s not just Windows workloads that enterprises are having a hard time modernising. “Windows is not the only legacy platform in the datacentre that is slowing down all your modernisation efforts … oftentimes it is VMware workloads that customers would really love to modernise to cloud-native services,” said Garman.

“VMware is deeply entrenched in many datacentres, and has been for a really long time. And what happens is … because it’s been there for a long time, there ends up [being] this kind of spaghetti mess of interconnected applications.”

“[So] really the hardest part about modernising is finding out what are the dependencies of those applications,” he said. “And the migrations are error-prone, because it’s hard to understand if you move something, if it is going to break something else. And again, of course, licensing is expensive.”

To assist with this, Q Developer also has capabilities that will allow VMware-based datacentre workloads to be reconfigured to become cloud-native, with the system able to identify the dependencies and create a migration plan for the user.

“[This] really reduces a ton of the migration time, and significantly it reduces [the organisation’s] risk,” said Garman. “It also launches agents that can convert on-premise VMware network configurations into modern AWS equivalents. This takes what used to be months and months of work into hours to weeks.”

The next complex datacentre migration project the company is looking to simplify for enterprises, with the help of Amazon Q, concerns mainframes, which Garman described as “by far the most difficult to migrate to the cloud”.

“When you talk to customers, just the effort of trying to analyse, document and plan mainframe modernisation is often too much, [and] people give up [because] it’s too hard. Turns out, Q can help with this, too,” he said.

The software has a number of agents in it that are able to do mainframe code analysis, refactor applications and create documentation in real time for legacy COBOL code so enterprises can fill in any knowledge gaps about what it might do.

“Most customers will tell you their mainframe migration will probably take three to five years … but planning a project for three to five years is nearly impossible,” said Garman. “A lot of the time, they just don’t get done.”

And while it’s beyond the capabilities of Amazon Q to make mainframe migrations a “one-click” job right now, he said early testing suggests the software could significantly accelerate the pace of these projects.

“We think Q can actually turn what was going to be a multi-year effort into a multi quarter effort, cutting by more than 50% the time to migrate mainframes,” said Garman. “If you can take a multi-year effort and bring it down to a couple of quarters, that’s something that people can really get their heads around. And customers are incredibly excited about this.”

Source

Posted on

2025: The year of AI for business – top trends to watch out for

You might not have started thinking about your Christmas shopping yet, but I bet you’ve been thinking about what artificial intelligence (AI) for business is going to look like in 2025. If you haven’t, then settle in with a glass of mulled wine, because now is your chance.

AI has come leaps and bounds over the past few years and is currently one of the biggest opportunities for business growth. With capabilities to intelligently automate admin tasks, take on customer service tasks, and analyse masses of data, the advantages are endless. But there’s still lots of room for development, in ways which will and won’t surprise you.

Stepping into the year of AI for business

Like your list of New Year’s Resolutions, the regulation landscape is constantly changing and adapting to the needs of tech businesses. For AI development to thrive in 2025, there must be a supportive environment ready for it. There’s no denying the appetite for AI, with over 120 bills on AI currently before the United States Congress. These build upon regulations already in place, such as the EU AI Act, which promotes the rapid adoption of trustworthy AI through reduced administrative burdens for SMEs and clear requirements for AI use.

The EU AI Act defines AI systems by their risk rating, splitting them up into prohibited, high-risk, limited-risk, and minimal-risk groups. This is something we could see changing in 2025, with the potential for new legislation focusing on AI classification over risk. This approach would consider criteria such as the intended uses and basic properties of AI systems.

New legislation coming into effect next year will significantly impact how businesses can use AI. Data management is one area likely to see substantial legislative focus, ensuring that AI does not compromise the security and privacy of business and customer data.

AI developments – The weird and the wonderful

As new legislation is rolled out in 2025, it will give businesses and developers more freedom and safety to largen AI’s scope. Many of us will already have AI ingrained into our processes, but what will we be bringing on board next?

  • Leading the way – Microsoft

One company which has been leading the way in AI development in 2024 has been tech giant Microsoft. At its recent Ignite 2024 event, it made several announcements which demonstrate the acceleration of AI in 2025. One of these was that Microsoft Teams will let participants speak in a language of their choice, through its new AI-powered Interpreter feature. Facilitating global communication and collaboration, this is one powerful way in which AI will fuel business growth.

Microsoft also announced the introduction of its AI agents this year. These agents will drive organisational wide optimisation and automation by collaborating with workers, a step forward from the AI assistants we already have. Agents can be trained to know your organisation from top to bottom and can compile details for business pitches and presentations whilst you focus on more valuable tasks.

  • Cutting corners with automation

Like AI agents, other AI systems which rely on trigger-based automation will flourish in 2025. Once the system is notified of a trigger, such as an email being received, it can digest the information and deliver an automated response to the trigger. Automated AI will seamlessly slot into business processes, taking care of admin tasks which frees up time for workers in all levels of the business to spend more time with customers and focus on their long-term needs.

The rise of automated AI poses a need for focus on responsible usage. Automation means that AI could be exposed to confidential data, and without the right protection measures in place, could learn that data and share it without authorisation. Legislation will play a key role in ensuring the responsible and ethical use of AI, but responsibility lies with business leaders as well to make sure that AI adoption goes hand in hand with education. Its important to understand that we will always include a human in the loop and full observability of these interactions with AI.

AI-powered systems might be forging new opportunities for businesses, but they lose their value and customer trust if inaccurate. To prioritise the accuracy of the models AI systems are trained on, we will see a shift in the New Year on how this process works. Grounding an model in accurate, secure data is extremely important. The better he data the more accurate the responses will be. Developers may synthesise their training data on large language models, and then train the AI system on a small language model.

This will approve the accuracy of the AI system, but as it adds degrees of complexity, it also poses the risk of potential bias or incorrect activity, such as the AI hallucination concept. When AI produces information like it is fact without any data to back it up, it’s a sign that something has gone wrong with the training data. Whilst 2025 will be a big year for the development of training models, businesses need to be aware of how their AI systems are being trained to avoid bias and unethical practice.

Not just a New Year’s Resolution

The huge amount of investment in 2025 is just one of many signs that AI isn’t a fleeting New Year’s Resolution. Companies like OpenAI and Microsoft have made a long-term commitment to investing in AI development, because they know we’re still unlocking its full portfolio of capabilities. Even if they’re not profiting off AI right now, it’s undoubtable that the future is rich. But this isn’t just a game for the big players, small businesses will also be staking their claim by adopting and investing in AI.

With the developments we’ll see next year in automation, robotics, and training data, it’s certain that there’ll be a flurry of businesses who haven’t explored AI yet looking to adopt. To make the most of the new developments, don’t wait until New Year’s Day to get started, reach out to the experts now to help your business get AI ready.

Chris Huntingford is the newly-promoted director of AI at ANS, a digital transformation provider and Microsoft’s UK Services Partner of the Year 2024. Headquartered in Manchester, it offers public and private cloud, security, business applications, low code, and data services to thousands of customers, from enterprise to SMB and public sector organisations.

Source

Posted on

It looks like the Nvidia App could be doing more harm than good by tanking game performance – but there’s a fix

The Nvidia App has been out for just over a month now – it’s Team Green’s replacement for GeForce Experience – but the new software is reportedly slowing down gaming performance for some folks.

A user named Sebastian Castellanos on X flagged up this issue, noting that with the Nvidia App installed, their gaming PC was being slowed down to the tune of 15%, and also suffering “horrendous” frametime issues (essentially jittery gameplay).

Castellanos said that this mainly happened with games using Unreal Engine 5 (specifically Black Myth: Wukong and The Talos Principle 2). Other X users also chimed in on the thread to note slowdowns to the tune of 10% to 15% or thereabouts.

First of all, Castellanos observed that the issues occurred with or without the Nvidia App’s overlay running in games, and subsequently posted again to say that they’d pinned down the problem to the ‘Game filters and Photo mode’ option in the app’s settings.

Apparently, turning this off cured the observed frame rate blues. Running some benchmarks with the game filters mode turned on, then off, Castellanos found that Black Myth: Wukong ran 22% faster with the mode disabled.

Stuttering was also way more pronounced with the mode enabled, as you can see in the graphs provided (check out the 1% percentile lows – the biggest dips in the frame rate) in the below post.

An Nvidia RTX 4070 Super on a purple deskmat on a desk

(Image credit: Future / John Loeffler)

Analysis: Hopefully this fudge will work for most gamers

As Tom’s Hardware reports – which also found similar levels of slowdown due to this bug – Nvidia has confirmed that it’s now investigating the purported glitch.

{ window.reliablePageLoad.then(() => { var componentContainer = document.querySelector(“#slice-container-newsletterForm-articleInbodyContent-cQNmmb6hAtbgKQDpdFxkB9”); if (componentContainer) { var data = {“layout”:”inbodyContent”,”header”:”Get daily insight, inspiration and deals in your inbox”,”tagline”:”Sign up for breaking news, reviews, opinion, top tech deals, and more.”,”formFooterText”:”By submitting your information you agree to the Terms & Conditions and Privacy Policy and are aged 16 or over.”,”successMessage”:{“body”:”Thank you for signing up. You will receive a confirmation email shortly.”},”failureMessage”:”There was a problem. Please refresh the page and try again.”,”method”:”POST”,”inputs”:[{“type”:”hidden”,”name”:”NAME”},{“type”:”email”,”name”:”MAIL”,”placeholder”:”Your Email Address”,”required”:true},{“type”:”hidden”,”name”:”NEWSLETTER_CODE”,”value”:”XTR-D”},{“type”:”hidden”,”name”:”LANG”,”value”:”EN”},{“type”:”hidden”,”name”:”SOURCE”,”value”:”60″},{“type”:”hidden”,”name”:”COUNTRY”},{“type”:”checkbox”,”name”:”CONTACT_OTHER_BRANDS”,”label”:{“text”:”Contact me with news and offers from other Future brands”}},{“type”:”checkbox”,”name”:”CONTACT_PARTNERS”,”label”:{“text”:”Receive email from us on behalf of our trusted partners or sponsors”}},{“type”:”submit”,”value”:”Sign me up”,”required”:true}],”endpoint”:”https://newsletter-subscribe.futureplc.com/v2/submission/submit”,”analytics”:[{“analyticsType”:”widgetViewed”}],”ariaLabels”:{}}; var triggerHydrate = function() { window.sliceComponents.newsletterForm.hydrate(data, componentContainer); } if (window.lazyObserveElement) { window.lazyObserveElement(componentContainer, triggerHydrate); } else { triggerHydrate(); } } }).catch(err => console.error(‘%c FTE ‘,’background: #9306F9; color: #ffffff’,’Hydration Script has failed for newsletterForm-articleInbodyContent-cQNmmb6hAtbgKQDpdFxkB9 Slice’, err)); }).catch(err => console.error(‘%c FTE ‘,’background: #9306F9; color: #ffffff’,’Externals script failed to load’, err)); ]]>

Sign up for breaking news, reviews, opinion, top tech deals, and more.

Team Green advises as follows: “We are aware of a reported performance issue related to Game Filters and are actively looking into it. You can turn off Game Filters from the Nvidia App Settings > Features > Overlay > Game Filters and Photo Mode, and then relaunch your game.”

It’s certainly worth trying that suggested fix, although Castellanos cautions that this might not work for everyone affected. If it doesn’t cure any sluggishness you might be experiencing with the Nvidia App, then about the only other option is to simply uninstall it and go without – until Team Green applies a fix. You can run with just the bare Nvidia graphics driver without needing the app, in case you were wondering.

Hopefully Nvidia will be able to swiftly implement a patch to resolve this one, as it’s a pretty nasty bug by all accounts – one that should arguably have been caught in the lengthy beta for the Nvidia App.

Via Wccftech

You might also like

Source

Posted on

Apple may be forced discontinue these 3 iPhone models in Europe

In September, Apple released the new iPhone 16 lineup. With that, the company finally stopped selling the 2021 iPhone 13 and only offered customers the following options through its store:

  • iPhone SE 3 (2022)
  • iPhone 14 (2022)
  • iPhone 14 Plus (2022)
  • iPhone 15 (2023)
  • iPhone 15 Plus (2023)
  • iPhone 16 series

However, the French Apple blog iGeneration reports that Apple will discontinue all Lightning iPhone models in Europe at the start of 2025.

Since European legislation requires all smartphones to have a universal USB-C port, the iPhone 14, iPhone 14 Plus, and iPhone SE 3 won’t follow the rule, as they have a Lightning port.

There’s always the possibility that Apple could tweak those smartphones with a USB-C port, as the company has done with the AirPods Max. However, it doesn’t seem Cupertino will follow that path.

Tech. Entertainment. Science. Your inbox.

Sign up for the most interesting tech & entertainment news out there.

By signing up, I agree to the Terms of Use and have reviewed the Privacy Notice.

Of course, a fourth-generation iPhone SE is expected to launch early next year, which means the current generation will soon be discontinued anyway. Besides that, when Apple introduces the iPhone 17 line, it will also stop selling the iPhone 14 and iPhone 14 Plus.

In other words, Apple is set to discontinue all three of these old iPhone models anyway in 2025. Still, if you’re not in Europe, you don’t need to worry about any of this, as Apple will continue to sell them until new models are introduced over the course of the year.

The new iPhone SE 4 is expected to feature a design like the iPhone 14 with a single camera, the A18 processor, and Apple’s new 5G modem. By the second half of 2025, Cupertino is rumored to launch the iPhone 17 lineup, including an all-new iPhone 17 Air, which will be the thinnest iPhone to date with some high-end features but hardware that’s not as good as an iPhone Pro.

BGR reached out to Apple but did not hear back at the time this story was published. We’ll update it if we hear from the company.

Source

Posted on

Nvidia might reveal DLSS 4 at CES 2025 – and mysterious new AI capabilities that could be ‘revolutionary’ for GPUs

  • Inno3D has leaked that Nvidia has “advanced DLSS technology” to show off at CES 2025
  • This may be DLSS 4, as it makes sense to reveal it alongside RTX 5000 GPUs
  • New neural rendering capabilities are also set to be aired which could be even more intriguing

Inno3D has again been leaking material relating to Nvidia’s upcoming revelations at CES 2025, but this time it’s more about the software and AI side of the equation, rather than the (purported) next-gen graphics cards themselves.

VideoCardz noticed that German tech site Hardware Luxx caught the CES 2025 press release from Inno3D, teasing what it has in store for the show, and oversharing some info that Nvidia would doubtless not want aired.

The key mentions here pertain to a possible new version of DLSS and fresh neural rendering capabilities.

In the first case, Inno3D talks about: “Advanced DLSS Technology: Nvidia’s Deep Learning Super Sampling offering even better image quality and higher frame rates.”

And secondly, the manufacturer points out: “Neural Rendering Capabilities: Revolutionizing how graphics are processed and displayed.”

There’s also talk of AI enhanced power-efficiency measures whereby the GPU’s power consumption and thermals are presumably fine-tuned to be more efficient and work better in general.

An Nvidia GeForce RTX 4060 Ti

(Image credit: Future / John Loeffler)

Analysis: Clever tricks to make up for meager VRAM loadouts?

While we can’t read too much into this – it’s all pretty vague marketing speak from Inno3D, as you’d fully expect from a pre-event press release – the highlighted bits are still exciting glimpses of what we might be treated to at CES 2025.

{ window.reliablePageLoad.then(() => { var componentContainer = document.querySelector(“#slice-container-newsletterForm-articleInbodyContent-Ep7bNnmUzbtxdMLDJ6tnSi”); if (componentContainer) { var data = {“layout”:”inbodyContent”,”header”:”Get daily insight, inspiration and deals in your inbox”,”tagline”:”Sign up for breaking news, reviews, opinion, top tech deals, and more.”,”formFooterText”:”By submitting your information you agree to the Terms & Conditions and Privacy Policy and are aged 16 or over.”,”successMessage”:{“body”:”Thank you for signing up. You will receive a confirmation email shortly.”},”failureMessage”:”There was a problem. Please refresh the page and try again.”,”method”:”POST”,”inputs”:[{“type”:”hidden”,”name”:”NAME”},{“type”:”email”,”name”:”MAIL”,”placeholder”:”Your Email Address”,”required”:true},{“type”:”hidden”,”name”:”NEWSLETTER_CODE”,”value”:”XTR-D”},{“type”:”hidden”,”name”:”LANG”,”value”:”EN”},{“type”:”hidden”,”name”:”SOURCE”,”value”:”60″},{“type”:”hidden”,”name”:”COUNTRY”},{“type”:”checkbox”,”name”:”CONTACT_OTHER_BRANDS”,”label”:{“text”:”Contact me with news and offers from other Future brands”}},{“type”:”checkbox”,”name”:”CONTACT_PARTNERS”,”label”:{“text”:”Receive email from us on behalf of our trusted partners or sponsors”}},{“type”:”submit”,”value”:”Sign me up”,”required”:true}],”endpoint”:”https://newsletter-subscribe.futureplc.com/v2/submission/submit”,”analytics”:[{“analyticsType”:”widgetViewed”}],”ariaLabels”:{}}; var triggerHydrate = function() { window.sliceComponents.newsletterForm.hydrate(data, componentContainer); } if (window.lazyObserveElement) { window.lazyObserveElement(componentContainer, triggerHydrate); } else { triggerHydrate(); } } }).catch(err => console.error(‘%c FTE ‘,’background: #9306F9; color: #ffffff’,’Hydration Script has failed for newsletterForm-articleInbodyContent-Ep7bNnmUzbtxdMLDJ6tnSi Slice’, err)); }).catch(err => console.error(‘%c FTE ‘,’background: #9306F9; color: #ffffff’,’Externals script failed to load’, err)); ]]>

Sign up for breaking news, reviews, opinion, top tech deals, and more.

The mentioned advanced DLSS tech which delivers a better image quality, and bigger frame rate boosts, might well be DLSS 4. That said, we’ve heard very little about Nvidia’s next-gen take on DLSS, which is odd if it is on the verge of being shown off.

However, it’s not unreasonable to assume that DLSS 4 would be tied to RTX 5000 GPUs exclusively (as Team Green did this with DLSS 3 and RTX 4000 GPUs when they launched). And so when RTX 5000 graphics cards are revealed at CES, it’d make sense that the next-gen DLSS would be teased alongside them, if not fully detailed.

On top of that, the apparent new neural rendering capabilities sound intriguing, and the mention of the term ‘revolutionizing’ graphics has piqued our curiosity. Is this just PR bluster, though?

We’ll have to wait and see, but there are already theories floating around that it could be some kind of neural texture compression, which would help GPUs with lower amounts of VRAM cope better with weighty textures. Could this be an explanation of why Nvidia might be mulling video RAM loadouts like 8GB for the RTX 5060 and 12GB for the RTX 5070? Perhaps, but that’s reaching…

Inno3D also mentions that it’ll have new graphics cards at CES 2025, without saying they’re RTX 5000 models. But it does mention some more standard brands of new products, alongside higher end iChill variants, including a small form-factor board – which is a hint that we won’t just see higher-end Blackwell GPUs at the show.

As well as the RTX 5090 and 5080, the RTX 5070 or 5070 Ti have been rumored as being ready to be revealed, and this is a further suggestion that this is what Nvidia has planned for CES in January.

Inno3D probably isn’t Nvidia’s favorite partner at the moment, because the graphics card manufacturer recently leaked the existence of the RTX 5090 and that it’ll be unveiled at CES 2025.

You might also like

Source