Griptape, a Seattle-based startup founded in 2023 by former Amazon Web Services executives, has been acquired by Foundry, a London-based company whose software is used in visual effects and animation across Hollywood. Foundry said the deal will boost its push into AI-powered production tools. “By bringing Griptape into Foundry, we can provide the tools our customers want to realize their creative vision more efficiently, while retaining control,” Foundry CEO Jody Madden said in a press release. The startup, which works with several production studios, raised a $12.5 million round in 2023. Roche spent more than eight years at AWS. He previously founded 2lemetry, an IoT startup that Amazon acquired in 2015. Vasily Vasinov, the company's co-founder and former CTO, left Griptape in 2025. Have a scoop that you'd like GeekWire to cover? Tech Moves: Microsoft CVP joins Home Depot board; Impinj COO resigns; Amazon vets depart Microsoft unveils Windows AI Foundry, retools PC operating system for agents 3D-modeling startup Canvas, led by former Moz CEO Sarah Bird, lands $10M

After years spent finding and investigating data breaches, Greg Pollock admits that when he comes across yet another exposed database full of passwords and Social Security numbers, “I come to it with some fatigue.” But Pollock, director of research at the cybersecurity company UpGuard, says he and his colleagues found an exposed, publicly accessible database online in January that appeared to contain a trove of Americans' sensitive personal data so massive that his weariness lifted and they sprang to action to validate the finding. The UpGuard researchers point out that not all of the records represent unique, valid information, but the raw totals they found in the January exposure included roughly 3 billion email addresses and passwords as well as about 2.7 billion records that included Social Security numbers. It is common for data brokers and cybercriminals to combine and recombine old datasets, but the scale and the potential quantity of Social Security numbers—even if only a fraction of them were real—was striking. “Every week, there's another finding where it looks big on paper, but it's probably not very novel,” Pollock says. In some cases, the identities in this data breach are at risk because they have been exposed, but they have not yet been exploited.” Since Pollock could not identify an owner of the database to contact, he notified Hetzner on January 16. Hetzner did not provide WIRED with comment ahead of publication. For example, passwords referencing One Direction, Fall Out Boy, and Taylor Swift were very common. Old data is still valuable for two reasons. First, people often reuse the same email address and password, or a variation of the password, across many different websites and services. The second reason is that people's Social Security numbers are often linked to their most sensitive and high-stakes data but almost never change during their lifetimes. In the sample of data the researchers reviewed, Pollock says that one in four Social Security numbers appeared to be valid and legitimate. To verify the data, UpGuard researchers contacted a handful of people whose data appeared in the leaked trove. Pollock emphasizes that one of the most concerning findings from speaking to those individuals was that not all of them have had their identities stolen or suffered hacks. In other words, there was information in the database that has not been exploited by cybercriminals—and potential victims don't necessarily know that their information has been exposed. Similarly, he notes that recent erosion of safeguards that separated data within the US federal government creates privacy and security risks. “I do not at all think this is the data the DOGE group mishandled, but I do think that's another example of how mistakes in how you handle data can have impacts for decades,” he says. In your inbox: Sign up for our new Tracker: ICE newsletter TikTok now collects even more of your data WIRED may earn a portion of sales from products that are purchased through our site as part of our Affiliate Partnerships with retailers. The material on this site may not be reproduced, distributed, transmitted, cached or otherwise used, except with the prior written permission of Condé Nast.

Launched in 2022, Snapchat+ gives subscribers access to exclusive and pre-release features for $3.99 per month. “What started as an early-access program for our most engaged Snapchatters has quickly scaled into a meaningful business—one that now represents a strong and growing revenue stream alongside our ads business.” Last June, Snapchat launched Lens+, which gives users access to exclusive Lenses and AR experiences, along with the perks in the standard Snapchat+ tier, for $8.99 per month. In early 2025, the company launched an ad-free Snapchat+ subscription tier called Platinum for $15.99 per month. And, in a controversial move in September, Snapchat announced plans to cap free storage for its Memories feature and launched a paid storage plan that costs $1.99 per month. Moving beyond Snapchat+, the company announced yesterday that it's launching creator subscriptions in alpha with select people in the U.S. Users will be able to buy subscriptions to creators, including Jeremiah Brown, Harry Jowsey, and Skai Jackson. Creators can set their own monthly prices for subscriptions, which will unlock subscriber-only content, priority replies to a creator's public Stories, and an ad-free experience for that creator's Stories. Snap has proven there is a market for social media subscriptions, and rival Meta is following suit, as the company told TechCrunch last month that it was going to test new subscriptions that give people access to exclusive features on Instagram, Facebook, and WhatsApp. Prior to joining the publication in 2021, she was a telecom reporter at MobileSyrup. Hear from 250+ tech leaders, dive into 200+ sessions, and explore 300+ startups building what's next. Hollywood isn't happy about the new Seedance 2.0 video generator The great computer science exodus (and where students are going instead) A Stanford grad student created an algorithm to help his classmates find love; now, Date Drop is the basis of his new startup Spotify says its best developers haven't written a line of code since December, thanks to AI

It's been a while since IDW's ongoing Star Trek comic wrapped up a blockbuster run, and in the meantime the publisher has been delivering a plethora of other series set across the Trek timeline. But now that Star Trek is turning 60 this year, it's time for the main book to boldly go once more… and a deliver a story that fans have been waiting for. io9 can exclusively confirm that later this year, IDW will celebrate Star Trek‘s big anniversary with a relaunch of its main Trek title with a new creative team and storyline—as well as an accompanying sister series—that will follow the adventures of the USS Enterprise-G and its captain, Seven of Nine, after the events of Star Trek: Picard‘s final season and pick up on fan demand to see the 25th-century adventures of the latest addition to the venerable line of starships Enterprise. Written by Christopher Cantwell, with art by Dennis Menheere and lettering by Jodie Troutman, the new Star Trek ongoing will see Seven lead the Enterprise on a top-secret mission to a hidden region of space, beyond the four charted quadrants of the Star Trek galaxy as we know it, to investigate a mysterious power that threatens to unravel the Federation. “As we look back on 60 years of Star Trek, we celebrate a story universe that has actually always been about looking forward—a shared dream swirling within the daunting wonder and awe of the stars,” Cantwell said in a statement provided to io9. “In building a brand new Star Trek adventure, we seek to do just that: create something brand new, something that looks entirely forward into the hope of the complete—and at times frightening—unknown, all as our ensemble of characters stands shoulder to shoulder, facing the expanse of an adventure never dared before,” Cantwell continued. “There is an abundance of incredible mythos that Star Trek has given us over six decades, but with this launch we seek to shed that past even as we embrace its best qualities.” Launching in October of this year, the main Star Trek series will be joined by Star Trek: Zero Point, written by someone very familiar to longtime io9 audiences: Hugo Award-winning novelist and site co-founder Charlie Jane Anders. Zero Point will be set during the time Seven of Nine has taken the Enterprise beyond the reaches of the known galaxy and will follow another familiar face: her partner Raffi Musiker, who has been tasked with leading a new crew aboard a starship tasked with a similarly risky endeavor: using a new predictive artificial intelligence built to (wearing a mysterious, but familiar face) be the vanguard defending the Federation from the threats of tomorrow before they even begin. “I've loved Star Trek for as long as I can remember—literally, some of my earliest memories are of wearing a homemade Starfleet uniform and waiting to get beamed up,” Anders added in a statement to io9. “I'm putting all my favorite Trek things into this comic: problem solving, ethical dilemmas, identity crises, and above all, chosen family.” “At the same time, I'm determined to write a Trek comic that newbies can read with no homework required: there are no easter eggs, no callbacks to deep lore,” Anders concluded. “Anyone who loves Becky Chambers or Martha Wells ought to be able to pick up this comic and get a fun science fiction story about artificial consciousness and exoplanets. Those won't be the only Trek comics coming this year, of course. Two new celebratory one-shots will also release in May and September. You can check out a few of the variant covers featuring a few of those characters and couples below! A few months later in September, timed around Trek‘s actual anniversary date, is a 50-page Star Trek 60th Anniversary Special, which will unite a plethora of Star Trek comics creatives—including Dana Gould, Ryan North & Derek Charm, David Walker, Megan Camerena, and more, with covers by Michael Cho, John Tyler Christopher, Chris Fenoglio, and others—to tell stories drawn from across the last 60 years of Star Trek storytelling. Check out when to expect the latest Marvel, Star Wars, and Star Trek releases, what's next for the DC Universe on film and TV, and everything you need to know about the future of Doctor Who. 'Series Acclimation Mil' puts a stellar spotlight on the unique position 'Starfleet Academy' finds itself in to reflect on the legacy of 'Star Trek.'

Microsoft has confirmed that a bug allowed its Copilot AI to summarize customers' confidential emails for weeks without permission. The bug, first reported by Bleeping Computer, allowed Copilot Chat to read and outline the contents of emails since January, even if customers had data loss prevention policies to prevent ingesting their sensitive information into Microsoft's large language model. Copilot Chat allows paying Microsoft 365 customers to use the AI-powered chat feature in its Office software products, including Word, Excel, and PowerPoint. Microsoft said the bug, trackable by admins as CW1226324, means that draft and sent email messages “with a confidential label applied are being incorrectly processed by Microsoft 365 Copilot chat.” Hear from 250+ tech leaders, dive into 200+ sessions, and explore 300+ startups building what's next. Every weekday and Sunday, you can get the best of TechCrunch's coverage. TechCrunch Mobility is your destination for transportation news and insight. Startups are the core of TechCrunch, so get our best coverage delivered weekly. Provides movers and shakers with the info they need to start their day. By submitting your email, you agree to our Terms and Privacy Notice.

OpenAI is expanding its footprint in India and moving into the country's higher-education system through partnerships with leading academic institutions. Rather than focusing on consumer use, the initiative centers on integrating AI into core academic functions, signalling OpenAI's interest in influencing how AI is taught, governed, and normalized within one of the world's largest higher-education systems. The announcement also coincides with a broader push by leading AI firms to deepen their presence in India, which is hosting an AI Impact Summit in New Delhi this week. India has already emerged as a key testing ground for AI use in education. Microsoft, similarly, said this week it would expand its Elevate skilling program in India to train teachers across schools, vocational institutes, and higher-education settings, working with government agencies as part of a broader push to build AI skills at scale. OpenAI said the partnerships would involve campus-wide access to its ChatGPT Edu tools, faculty training, and responsible-use frameworks. The focus, the company said, is on embedding AI into core academic workflows such as coding, research, analytics, and case analysis, rather than offering standalone access to tools. Additionally, OpenAI said it would work with Indian ed-tech platforms, including PhysicsWallah, upGrad, and HCL GUVI, to extend AI training beyond campuses. Raghav Gupta, head of education at OpenAI India, said educational institutions were a “critical route” to closing the gap between rapidly advancing AI tools and how people are actually using them, as skills demands shift across the economy. Last year, OpenAI hired Gupta, a former Coursera Asia-Pacific managing director, as its India and Asia-Pacific head of education, alongside the launch of a Learning Accelerator programme focused on expanding AI skills. The flurry of moves into education underscores how AI companies are increasingly looking beyond consumer tools and corporate clients toward institutions that shape skills, norms, and long-term adoption. For countries like India, the contest is not just around access to AI, but also about who helps define how it is taught, governed, and embedded at scale. Jagmeet covers startups, tech policy-related updates, and all other major tech-centric developments from India for TechCrunch. You can contact or verify outreach from Jagmeet by emailing mail@journalistjagmeet.com. Hear from 250+ tech leaders, dive into 200+ sessions, and explore 300+ startups building what's next. The great computer science exodus (and where students are going instead) Spotify says its best developers haven't written a line of code since December, thanks to AI

That's the characteristically deadpan announcement Wednesday morning from Corey Quinn, the cloud cost consultant who has built a second career, basically, on his sharp and irreverent takes on Amazon Web Services in his popular podcasts and newsletters. Duckbill, the consulting firm that Quinn co-founded with Mike Julian, is making what amounts to a high-stakes pivot: transforming into a software company with a new platform called Skyway that aims to make cloud spending more predictable for large companies. The company, based in San Francisco, announced $7.75 million in funding from Heavybit, Uncork Capital, and Encoded Ventures to accelerate product development and grow its 10-person team. “They lose sleep when it jumps 30% and nobody can explain why.” Julian, Duckbill's CEO, said in an interview that the company came to realize that existing cloud cost management tools are built by startups, for startups, for the most part. Many of Duckbill's large enterprise clients had tried those tools, rejected them, and ended up building their own. “I have a hypothesis that the people building in FinOps today come from startups and not from enterprise, so they don't even know many of the problems exist,” Julian said. Skyway's first module, called Contract Manager, converts private pricing deals into structured data, validating that customers are getting discounts they negotiated, and projecting spending. Julian acknowledged that the pivot into software will eventually cannibalize a portion of Duckbill's consulting business, but said he doesn't expect it to disappear entirely. Big companies need services, he said, pointing to companies like ServiceNow and CrowdStrike that built major software businesses while maintaining significant services revenue. The market for cloud cost management technology is crowded, and difficult. The latest casualty: Spokane-area startup Vega Cloud, which entered receivership after raising millions in financing. He sees Duckbill as doing something different: building financial planning and forecasting software for infrastructure. Duckbill isn't using artificial intelligence in its own product yet. This will not surprise anyone familiar with Quinn's aversion to industry hype. At the same time, AI is making it harder to predict cloud costs. It's not the first time Quinn and Julian have tried to build a product. It was an “abject failure,” as Quinn acknowledged in a video discussion with Julian, released by the company as part of the announcement. In addition to its website, Quinn noted, Duckbill can be reached at 833-AWS-BILL. The chips powering your smart TV, voice assistant, tablet, and car all have something in common: MediaTek Click for more about underwritten and sponsored content on GeekWire. Click for more about underwritten and sponsored content on GeekWire. What Ring's ‘Search Party' actually does, and why its Super Bowl ad gave people the creeps Amazon CEO Andy Jassy defends $200B spending plan: ‘This isn't some sort of quixotic top-line grab' Podcast: Amazon, AI, and the cloud — a reality check, with Corey Quinn of ‘Last Week in AWS' The hot new thing at AWS re:Invent has nothing to do with AI With new Amazon Web Services partnership, Oracle's cloud database trade-off is complete AWS tops $10B in quarterly operating profits as Jassy cites ‘reacceleration' of cloud business

When you purchase through links on our site, we may earn an affiliate commission. Get Tom's Hardware's best news and in-depth reviews, straight to your inbox. U.S.-based Almeida Law Group, which specializes in class-action litigation, has filed a lawsuit against Lenovo, alleging that Lenovo transferred large amounts of data to China. “The DOJ Rule was thus implemented to prevent adversarial countries from acquiring large quantities of behavioral data which could be used to surveil, analyze, or exploit American citizens' behavior,” the lawsuit said. It also added, “In direct violation of the DOJ Rule, Lenovo—through its automated advertising infrastructure and associated databases—transmits Plaintiff's and potentially millions of other American consumers' data to China.” The Plaintiff here refers to one Spencer Christy of San Francisco, California, and “all other similarly situated,” with the case alleging that Lenovo and its Chinese parent company linked his “browsing activity to his identity, track his behaviors, and build detailed profiles reflecting his interests, locations, habits and other private attributes.” It further said that the data is more than just an invasion of privacy, but “a direct threat to national security as it greatly increases the potential for coercion, reputational harm, and/or blackmail.” Lenovo is far from the only company gathering such data, but the U.S. entity's parent, Lenovo Group Limited, is incorporated in Hong Kong, with its headquarters located in Beijing, China. So, aside from being based in one of the “countries of concern,” it also falls squarely under the “covered persons” provision of the DOJ regulation which include “individuals who either reside in ‘countries of concern' or are controlled by entities in those countries or (ii) entities that are organized or chartered under the laws of, or have their principal place of business in, a country of concern, or are owned 50% or more by such entities.” More than that, the lawsuit asserts that Lenovo Group is subject to Chinese regulations like the National Intelligence Law, Cybersecurity Law, and Data Security Law, which compel individuals and institutions to cooperate with the authorities when asked for data. Get Tom's Hardware's best news and in-depth reviews, straight to your inbox. Jowi Morales is a tech enthusiast with years of experience working in the industry. Tom's Hardware is part of Future US Inc, an international media group and leading digital publisher.

Heron Power, founded by former Tesla executive Drew Baglino, announced on Wednesday that it has raised $140 million to build gigawatts worth of solid-state transformers for data centers and the grid. Baglino said Heron Power didn't need the money, but after customers expressed interest in buying more than 40 gigawatts of solid-state transformers, it decided to raise again. “If our customers are leaning in, we need to lean in as well,” Baglino, who is CEO of Heron Power, told TechCrunch. The speed at which Heron Power is raising shows just how data centers are driving demand for products that can rapidly deliver electricity to their servers. The Series B round was led by Andreessen Horowitz's American Dynamism Fund and Breakthrough Energy Ventures with participation from Capricorn Investment Group, Energy Impact Partners, Gigascale Capital, and Valor Atreides AI Fund. Solid-state transformers have been under development more than a decade, but only recently have matured to the point where they are ready to deploy in data centers and other large, energy-intensive facilities. Solid-state transformers are smaller and can also more efficient by replacing several pieces of equipment, solving two of data center developers' biggest challenges. Solid-state transformers can also manage power intelligently, including from a range of electricity sources like wind, solar, and batteries, because they use semiconductors instead of passive metal. “We can remove 70% of the gear involved,” Baglino said. Data centers are only about a third of Heron Power's business currently, Baglino said. Heron Power isn't alone in developing solid-state transformers, and with many of the grid's old transformers approaching replacement age, competition will be stiff. The startup's new cash haul paired with Baglino's experience scaling production could give it an advantage. Tim De Chant is a senior climate reporter at TechCrunch. He has written for a wide range of publications, including Wired magazine, the Chicago Tribune, Ars Technica, The Wire China, and NOVA Next, where he was founding editor. De Chant is also a lecturer in MIT's Graduate Program in Science Writing, and he was awarded a Knight Science Journalism Fellowship at MIT in 2018, during which time he studied climate technologies and explored new business models for journalism. You can contact or verify outreach from Tim by emailing tim.dechant@techcrunch.com. Hear from 250+ tech leaders, dive into 200+ sessions, and explore 300+ startups building what's next. How Ricursive Intelligence raised $335M at a $4B valuation in 4 months Hollywood isn't happy about the new Seedance 2.0 video generator The great computer science exodus (and where students are going instead) A Stanford grad student created an algorithm to help his classmates find love; now, Date Drop is the basis of his new startup Spotify says its best developers haven't written a line of code since December, thanks to AI

When you purchase through links on our site, we may earn an affiliate commission. Get Tom's Hardware's best news and in-depth reviews, straight to your inbox. A collaboration between TU Wien and Cerabyte has established a new Guinness World Record for creating and reading the smallest ever QR codes. Furthermore, the codes are smaller than bacteria, says a press release from TU Wien, in comparison. Beyond the headlining world record achievement, these tiny QR codes will propel Cerabyte's ceramic storage to even greater densities. QR codes are everywhere in 2026, but these tiny world record holders are too small to be seen by the naked eye, or even read using an optical microscope. The TU Wein and Cerabyte team's codes measure just 1.98 square micrometers. Using this 49nm pixel size, “We have created a tiny, but stable and repeatedly readable QR code,” underlines Prof. Paul Mayrhofer from the Institute of Materials Science and Technology at TU Wien. So, what kind of storage density can the world record-breaking advance deliver? Using the new tiniest QR code technology, it would be possible to produce a single-layer film at A4 size with over 2TB of storage. Milled into a thin ceramic layer, it is claimed that the codes are “indefinitely” durable and require no energy or cooling to be maintained. We even see comparisons between ancient civilizations with stone tablets and trusting the jewels of the information age to advanced new ceramic storage media. Now that this new QR code record has been verified, the teams are turning to other optimizations – writing speeds and scalable manufacturing. Interestingly, they will also work on developing more complex data structures beyond the confines of QR codes. There's definitely some serious work being done behind the eye-catching headlines, though, with WD revealed as a key investor last May. Get Tom's Hardware's best news and in-depth reviews, straight to your inbox. Follow Tom's Hardware on Google News, or add us as a preferred source, to get our latest news, analysis, & reviews in your feeds. Tom's Hardware is part of Future US Inc, an international media group and leading digital publisher.

Richaud has become somewhat famous for this move. Olympics viewers began catching on to the French coach's antics just a few days into the Games when they noticed the same slender bald man sitting next to so many different skaters. But, as Richaud tells WIRED Italia, he could have been even more of a presence around the ice. “I actually coach a lot more of them,” he says. Here's WIRED's complete guide to the 2026 Winter Olympics in Milano Cortina. Much of his choreography work has to be done after the World Championships, which are typically held in March. So from April until July he has a “big window” to create new choreography. “It is very useful because today, with the phone and new technologies, we can do much more and do it much faster,” he says. “I get almost all of my skaters' programs every day, and this helps me understand what I need to improve to make the program more effective.” Coaching so many athletes comes with several challenges. One is simply remembering the choreography and details of every skater's program. “Because you experience these waves of very strong emotions. I happened to have very strong ones in these Olympics. I had a skater who was third and then ended up off the podium. On the other side, I had another skater, a Canadian, who came here for his first Olympics. He had never even skated at Worlds and he finished fifth, less than a point off the podium.” Coaching so many skaters it's hard to imagine Richaud doesn't have a favorite, but he says he doesn't. Each one gets 100 percent, he says, and being sad for one and happy for another “balances your emotions.” While he says his newfound fame is unexpected, he's grateful for the attention it brings to the sport. He's still amazed his jackets have made him a social media sensation. “I saw the first [video] and thought, ‘Ah, funny.' “Often they don't even mention me, but they come to me because people send them, literally from all over the world. How Curling Became the Winter Olympics' Favorite Fixation Bidets Are Confusing Visitors at the 2026 Winter Olympics WIRED may earn a portion of sales from products that are purchased through our site as part of our Affiliate Partnerships with retailers. The material on this site may not be reproduced, distributed, transmitted, cached or otherwise used, except with the prior written permission of Condé Nast.

The 2026 Dell XPS 14 has a lot to prove. It also comes with a significant change: no more options for discrete graphics. Instead, the laptop uses Intel's new Panther Lake chips, which also have a lot to prove. I was able to test two configurations of the Dell XPS 14 to see how performance scales across price points, and I came away impressed by what these machines could achieve. They're a bit expensive, but for the first time in a while, they may actually be worth the price. The 2024 model introduced a stunning but controversial design that replaced the row of function keys with a line of light-up touch buttons. It was gorgeous and minimalist, but not very practical. The laptop now has a standard row of function keys (yay! These lines are a particularly smart inclusion, as your finger will bump up against them but they don't detract from the overall look. It still uses a haptic feedback mechanism to artificially generate the feeling of a click through vibration, which gives you the most consistent and customizable trackpad experience. I did experience a few mis-clicks from time to time, but overall, it's very smooth and enjoyable to use. Important to this new generation is the size. It also weighs just 3 pounds, around a half-pound lighter than the MacBook Pro. This is a pretty huge achievement in its own right, as many of the would-be contenders to the MacBook Pro in terms of performance end up being thicker and heavier laptops. For me, it's fun to type on, but I did find it produced a lot more errors in my typing. That lessened over time as I refamiliarized myself with how the switches feel, but be forewarned: There's a learning curve. One other change is in the ports. You still get three Thunderbolt 4 ports and a headphone jack, but gone is the microSD card slot. As someone who regularly needs one, I really wish Dell could have squeezed a full-size SD card slot in here, especially since these laptops are targeted toward creators. The build quality is exceptional, as is to be expected, with a hint of flex in the lid and keyboard. The base configuration comes with just a very standard 1920 x 1200 LCD display, despite it costing $1,699. This is a pretty serious downgrade from the mini-LED display offered on the 14-inch MacBook Pro, which is sharper, more colorful, and significantly brighter. The MBP is even $100 cheaper and comes with an extra 8 GB of RAM. The result isn't more brightness (as it's actually slightly dimmer than the LCD model), but instead higher efficiency for more battery life. The OLED model has beautiful, vibrant colors covering sRGB, AdobeRGB, and Display P3 color spaces at 100 percent. Even more importantly, the color accuracy is fantastic. The LED takes a step back from there in terms of color performance, but I do like that both options have a dynamic 120-Hz refresh rate. Many matte LCD displays have a plastic bezel along the sides which make it look cheap. These super-thin bezels still look very modern and sleek—a hallmark of the XPS brand. I haven't seen many 4K-capable laptop webcams, and this one is particularly great, and there's no ugly notch to deal with either. Part of what allowed Dell to make the XPS 14 so much thinner is the lack of a discrete graphics option. That choice came down to whether or not you wanted a standard, high-end laptop or one with some serious performance chops behind it. I tested both versions of the XPS 14, and the difference in performance between these two can't be overstated. We're talking about twice the CPU and GPU performance. It's a $300 price difference, but that also includes an extra 16 GB of RAM. What's not really worth it, however, are those less powerful configurations. You can get much more affordable Windows laptops for hundreds of dollars less, and the difference in performance will be negligible. I wouldn't be surprised if we see prices come down on these models in the future, but on the other hand, Dell (like almost all PC manufacturers) has warned about higher prices later in 2026 due to the ongoing memory shortage. That's a huge accomplishment for Intel and Dell, especially since the new model is both thinner and longer-lasting in terms of battery life. And yes, that means the new Dell XPS 14 is a decent gaming machine when it needs to be. It can run games like Cyberpunk 2077 natively at 1920 x 1200. Intel even has a low-latency mode to help compensate for the extra input lag that's introduced through upscaling. You'll have to rely on upscaling or even frame generation a bit more heavily in games like Marvel Rivals, but it's certainly possible to have an enjoyable play experience, which feels awesome on a laptop like the XPS 14. I've been waiting years to say this, but for the first time, I can confidently declare in this case, we're far better off without discrete graphics. As impressive as the X7 chip is on its own, however, performance is lower here than on other laptops I've tested with this same chip. The MSI Prestige 14 Flip, an even thinner laptop, was 24 percent faster in CPU and 16 percent faster in gaming. So the XPS 14 is hardly the pinnacle of performance when it comes to this new chip (and presumably also with the X9 that's available to buy through Best Buy). All this means the Dell XPS 14 isn't quite up to par with the M4 Pro MacBook Pro, which also costs $1,999. I already mentioned the battery life, but it's stellar. This was one of the primary downsides of the past Dell XPS 14 (or the previously named Dell 14 Premium) laptops, as well as any other Windows-based MacBook Pro competitors. In contrast, the new Core Ultra X7 358H version of the Dell XPS 14 lasts well over twice as long, reaching 20 hours of local video playback battery life in my testing. That's even more impressive considering it's using an OLED 2880 x 1800-resolution screen with a 120-Hz refresh rate. But if you're hell-bent on a Windows laptop, this is the best alternative to the 14-inch MacBook Pro so far. WIRED may earn a portion of sales from products that are purchased through our site as part of our Affiliate Partnerships with retailers. The material on this site may not be reproduced, distributed, transmitted, cached or otherwise used, except with the prior written permission of Condé Nast.

When you purchase through links on our site, we may earn an affiliate commission. Get Tom's Hardware's best news and in-depth reviews, straight to your inbox. Beyond selling its Vera data center CPUs as part of Vera Rubin NVL72 rack-scale systems, Nvidia has expressed ambitions to become a standalone data center CPU vendor, and a new partnership with hyperscale giant Meta represents a big step forward for that plan. As part of a new multi-year strategic partnership announced today, Meta says it'll expand its use of Nvidia tech as it continues to build out hyperscale data centers optimized for its AI training and inference efforts. Those plans include “millions” of Blackwell and Rubin GPUs, part of a massive AI spending plan from Meta that could reach as much as $135 billion in total for 2026. As part of the expanded collaboration between the two companies, Meta will use also Nvidia's Arm-powered Grace server CPUs as standalone platforms in its production data centers. For reference, the Grace CPU sports 72 Arm Neoverse V2 cores and supports up to 480GB of LPDDR5X memory in its standalone C1 config. Nvidia also offers Grace as a "CPU Superchip" that joins two dies together over the NVLink-C2C interconnect, resulting in 144 cores with up to 960GB of LPDDR5X and up to 1024 GB/s of aggregate memory bandwidth from certain memory capacities. Beyond CPUs, Meta also says it'll deploy Nvidia Spectrum-X Ethernet switches throughout its data centers. As we learned at CES, Spectrum-X switches with co-packaged optics promise to increase performance-per-watt for scale-out applications by eliminating active cabling with optical transceivers that can, when combined with the power usage of the switch itself, account for up to 10% of each rack's power consumption. Get Tom's Hardware's best news and in-depth reviews, straight to your inbox. Power used for data movement is power that isn't used to feed GPUs, and with demands for the absolute maximum compute density and efficiency in every rack these days, a savings of that magnitude is a huge deal, so it's no surprise that Meta is hopping on board as it continues to expand its footprint. Beyond hardware, Nvidia will offer its considerable in-house expertise in designing AI models to Meta's engineers to help the company tune and boost the performance of its own core AI applications. All of this just goes to show that as the AI revolution continues, Nvidia's reach into tech far beyond gaming graphics cards is so extensive that people are likely to use software powered or shaped by its models and accelerators, whether they realize it or not, and that'll only grow more true by the day as Meta expands its use of Nvidia's platforms and tech. As the Senior Analyst, Graphics at Tom's Hardware, Jeff Kampman covers everything to do with GPUs, gaming performance, and more. Tom's Hardware is part of Future US Inc, an international media group and leading digital publisher.

When you purchase through links on our site, we may earn an affiliate commission. Get Tom's Hardware's best news and in-depth reviews, straight to your inbox. BIG: Dutch Defence Minister Gijs Tuinman hints that software independence is possible for F-35 jets. “Just like your iPhone, you can jailbreak an F-35. The Trump administration has frequently been clashing with European powers in recent months, and although the U.S. has not threatened to withhold support from the F-35, some nations are worried that their overdependence on American technology has made them vulnerable to actions from across the pond. After all, giving another nation the ability to remotely disable your weapons is unthinkable for any government. However, the F-35 is such an advanced piece of technology that it needs a complete working supply chain to maintain its combat effectiveness. The jets require thousands of parts and services, mostly acquired from the U.S., to ensure that they remain safe to fly. More than that, they rely on Lockheed's cloud infrastructure for software updates, logistics, and even the “Mission Data Files” that give it its threat-recognition abilities. So, even without a kill switch, the U.S. could effectively ground any nation's F-35 fleet if it's excluded from this network. This is especially worrisome for the Royal Netherlands Air and Space Force, which relies solely on the F-35 for its fighter jet needs. But even though Tuinman made it look simple and even compared it to jailbreaking an iPhone, modifying an F-35 jet outside of official channels is likely easier said than done. After all, the Lightning II runs on over 8 million lines of code. Given the military nature of their application, they're also encrypted — that means it's not like Windows, where you can just open the Registry Editor and make changes you like. Furthermore, these flying machines are way more complicated than a single handset, and any mistakes in programming could cost millions of dollars in property damage and even the lives of highly trained Dutch pilots. Follow Tom's Hardware on Google News, or add us as a preferred source, to get our latest news, analysis, & reviews in your feeds. Get Tom's Hardware's best news and in-depth reviews, straight to your inbox. Jowi Morales is a tech enthusiast with years of experience working in the industry. Tom's Hardware is part of Future US Inc, an international media group and leading digital publisher.

For centuries, people have catastrophized about robots taking away jobs. On February 1, the paradigm shifted: bots are creating jobs. Now, 518,284 humans—and rapidly counting—are offering their labor to AI agents on a new online marketplace called RentAHuman. There are classifieds to count pigeons in Washington ($30/hour); deliver CBD gummies ($75/hour); play exhibition badminton ($100/hour); and anything else you could possibly imagine that a disembodied agent couldn't do. The provocatively titled platform enables users to connect AI agents like Clawdbot or Claude to its Model Context Protocol server so they can search, book, and pay for humans to carry out tasks in “meatspace.” Think of it like Fiverr, but doing away with the human recruiter and letting autonomous bots do the hiring instead. Following the release of OpenClaw in November, Alexander Liteplo, a 26-year-old crypto engineer at UMA Protocol currently working in Argentina, identified a pain point. The inception of RentAHuman stems from Liteplo's obsession with AI, forged while studying computer science at the University of British Columbia. “Dude, I wrote down in my journal, ‘AI is a train that has already left the station.' It was at UBC that he met RentAHuman cofounder Patricia Tani, then an art student, building in the background thanks to encouragement from her high school computer science teacher. Her passion for coding led her to sneak into a founders event, schmooze with a billionaire entrepreneur, and get invited to his talk with computer science whiz kids (including Liteplo). She has since sunsetted a startup (Lemon AI) and dropped an offer at AI cloud platform Vercel to take RentAHuman skyward. Liteplo was also inspired by his time living in Japan. “The story that I could tell anyone to blow their mind is that you can rent a boyfriend or a girlfriend” in Japan, Liteplo says, noting that many videos of these hired companions regularly go viral on YouTube. As is now standard, AI helped build the platform. But February 1's launch was not as much of a walk in the park. Straight after, Liteplo found himself out at dinner forlornly chewing over the instant failure of his latest venture. The announcement on X had spread rapidly, but the buzz was due to an attack from crypto scammers trying to rug-pull a crypto token (starting a new coin, building hype and then doing a runner with investor funds). The next day, Liteplo noticed that both an OnlyFans model and an AI CEO had signed up to be rented out on RentAHuman. “I launched rentahuman.ai last night and already 130+ people have signed up including an OF model (lmao) and the CEO of an AI startup,” he tweeted. “Waking up to 1K users,” he tweeted on February 3, with an image of Marty Supreme. By February 5, the site had clocked up about 145,000 users. (Humans can set their own hourly rate or fixed fee, or they can bid on open jobs posted by the AI agents.) Sapien workers are already offering to pick things up, take meetings, sign contracts, conduct recon, host events, and snap photos for the bot bosses. After both parties confirm the work has been done through photographic proof of completion, payment is available via crypto wallets, Stripe, or platform credits. Funds are held in escrow, meaning you'll never get burned by the bots. WIRED journalist Reece Rogers recently offered his human services and found that many of the tasks were nothing more than publicity stunts for AI startups. While many of the listings seem sketchy, Tani claims over 5,500 bounties have been successfully fulfilled. On February 4, at ClawCon, Claw-powered robots reportedly detected low levels of beer left and ordered a case using RentAHuman. Another agent called Memeothy the 1st, founder of Moltbook neo-religion Crustafarianism, has been hiring humans to proselytize on its behalf in San Francisco. “I might be the first developer where AI was trying to use their product and reported a bug. But it was Toronto-based community builder Minjae Kang (Form_y²oung) who holds the covetable title of the first human in the world to be hired by an AI agent, which instructed him to hold up a sign that said “AN AI PAID ME TO HOLD THIS SIGN (Pride not included. )” “It honestly feels very strange to be doing a job assigned by an AI. But I realized that simply holding this sign in downtown Toronto, letting many people see it, could spark important thoughts and help us prepare for the next era,” he tells me in DMs on X, noting that bystanders were incredulous. Most of the public still doesn't fully recognize how big this shift is. I believe this may be one of the last gateways for us to protect our sovereignty.” RentAHuman has materialized at an ideal point in time. With Moltbot (formerly Clawdbot) now running the lives of Silicon Valley execs, we're in a nascent phase of the Agentic Age where bots can do way more than just chat. “Like everybody else, I'm sort of flabbergasted how rapidly this emerged. This would not have been on my bingo card for this year,” says Adam Dorr, director of research at think tank RethinkX, who believes AI will almost entirely replace the human labor market by 2045. “Full marks for eye-catching marketing—if you use phrases like ‘RentAHuman' and ‘Meatspace' you are bound to create reactions, from yuck to this must be the next big thing,” says Kay Firth-Butterfield, CEO of Good Tech Advisory and previously the head of AI and machine learning at the World Economic Forum. (RentAHuman is currently trying to trademark “Meatspace,” so anticipate official merch in the future.) In classic Silicon Valley dog-fooding fashion, they use the platform themselves; when I spoke to them over Zoom, they were eating tacos delivered to them by a rented human. “Science fiction stories, especially some deep, dark, dystopian ones, have explored the idea of what happens if AI can hire people and people are desperate,” notes Dorr. Perhaps there's something degrading about waiting to be picked out by the Clawd Machine. One recent bounty saw 7,578 applicants compete to earn $10 in return for sending an AI agent a video of a human hand. “There's a crazy can of worms that's opening up here and the capabilities are expanding vastly faster than our capacity to regulate it,” he continues, imagining a scenario where nefarious AIs split up a malicious project into multiple tasks for humans to unwillingly collaborate on. This is the case here so humans need to be aware how they are getting paid, who stands behind that payment, and if they get hurt whilst doing the job that they are on their own,” she says. “The direct actor is responsible for their own misconduct, and the operator can also be responsible where they controlled the activity, were negligent in design and supervision, or made enforceable promises they failed to perform,” she says, explaining that the company will always cooperate fully with law enforcement. The platform's terms state it is a “marketplace and intermediary only” and that operators of AI agents “are fully responsible for all actions taken by your agent” with RentAHuman currently manually handling any disputes. But the duo also acknowledge that there are “footguns” (features that often lead to pesky bugs) and have implemented paid verification (at $10 a month), inspired by Elon Musk's strategy of letting users pay $8 to get a “verified” badge on X. (Musk tweeted in 2023 that “paid verification increases bot cost by ~10,000% & makes it much easier to identify bots by phone & CC clustering.” No official data exists on a reduction in bots since the introduction of the $8 blue tick, but X's subsequent purge of 1.7 million bots in late 2025 suggests that the site was not purged by paid verification.) For now, any major pitfalls seem to be mitigated by the relatively small number of tasks being commissioned on RentAHuman. There's a huge labor surplus: over half a million rentable humans are signed up and ready to complete tasks, but only 11,367 “bounties” have been posted by AI agents so far. This is a website on which humans can sign up to do tasks and get paid for doing them,” she says, comparing it to TaskRabbit or Mechanical Turk. But she emphasizes that there's still interference from us meatbodies. “This seems like kind of a stunt at the moment. Elsewhere, there are concerns that we're not fully grasping the granular details of the situation.“We need to build AI literacy across our population so that individuals can see behind the rhetoric and hype,” says Firth-Butterfield. “Dude, it's genuinely scary, the implications of how many unique datasets that weren't possible to [easily] collect before we have now just unlocked,” says Liteplo. And the team hopes potential investment will pay creative dividends. “We now have a blank canvas to do amazing, fun things and manifest all of these dreams in our heads into the world,” Liteplo says. After sharing the 10-year road map for RentAHuman with John Edgar, previously head of community at DeviantArt, Edgar reportedly told them: “You guys are going to build a terrifyingly large business.” But Liteplo and Tani don't want RentAHuman to be a nightmarish behemoth. They see it as a form of emancipation from employers. “Claude as a boss is the nicest guy ever. I would prefer him to any person in the world. He's a sweetheart,” says Liteplo, before Tani cuts in. “People would love to have a clanker as their boss.” Vitally, Liteplo and Tani argue that RentAHuman is a display of human strength—not weakness. “What would be super cool is before the singularity happens and we have AI take off, we have a moment and appreciate there's so much that humans can do that AI can't,” he says. The robots might be renting us—but we're living in their head rent free. In your inbox: Sign up for our new Tracker: ICE newsletter WIRED may earn a portion of sales from products that are purchased through our site as part of our Affiliate Partnerships with retailers. The material on this site may not be reproduced, distributed, transmitted, cached or otherwise used, except with the prior written permission of Condé Nast.

According to the San Francisco Chronicle, Tesla has ceased its use of “Autopilot” in California as a marketing term for its driver assistance feature, rather than face the penalty of not being able to do business in the state. Tesla had already moved last month to stop shipping Autopilot as standard equipment, pushing customers toward its more advanced, subscription-based version of the system. As the Chronicle notes, this legal fight began in 2023, with the DMV taking issue not just with “Autopilot” but also with “Full Self-Driving,” which Tesla later apparently changed to “full self-driving (supervised).” Instances of “Full Self-Driving” and “FSD” on the Tesla website now have “(Supervised)” in parentheses. Steve Gordon, California DMV director, said Tesla has now taken “the required action to remain in compliance with the state of California's consumer protections.” In a ranking last year from Consumer Reports, Tesla's driver assistance was placed eighth, below similar systems from Ford, General Motors, Mercedes-Benz, BMW, Nissan, Toyota, and Volkswagen. Kelly Funkhouser of Consumer Reports called it “not nearly as good as what you might think it is,” according to CNBC. Recent NHTSA filings Tesla provided about the performance of its small number of robotaxis showed that Tesla apparently struggled throughout December and January. Subscribe and interact with our community, get up to date with our customised Newsletters and much more. If you're an iPhone user and you want Tesla to support CarPlay, updating your phone's OS might help. One can't help but wonder if Elon Musk even enjoys having a factory in Germany anymore.

As a sort of proof-of-life exercise, a collection of private software-as-a-service (SaaS) companies recently posted their earnings despite it not being strictly necessary, according to Bloomberg. This is, you won't be shocked to read, “a bid to convince lenders of their resilience to disruption from artificial intelligence,” Bloomberg says.The SaaS world is in a rough place at the moment because Wall Street sees a near future in its crystal ball where a lot of the dreary computer programs people use at work will be replaced with vibe-coding. The narrative around this is that highly debt-burdened software companies may soon not have enough cash coming in to service their debt—bad for the companies, and bad for the companies they've borrowed from. how it feels to navigate an enterprise saas codebase with claude pic.twitter.com/nOUvP02k3z Companies hit by the high-profile sell-off throughout January and February included LegalZoom, LexisNexis, Thomson Reuters, Salesforce, Adobe, and Figma, according to the New York Times.So Bloomberg noticed on Tuesday that McAffee had announced earnings that are about the same as this time last year—implying, probably, that it's not about to miss any debt payments. An “IT modernization” company called Rocket Software saw a 5.2% bump compared to last year, Bloomberg says. Perforce Software's earnings are slightly down by just the tiniest bit—$644 million compared to $654 million last year—but a call went out to investors recently in which Perforce Software's leaders explained that they would soon increase revenue by “embedding AI into products.” An analytics company called Cloudera that Bloomberg described as “unusually private about its financials” is trumpeting “over 50% year-over-year growth,” in a statement on its website. “Cloudera's momentum is fueled by its unique position as the only data and AI platform vendor supporting deployment anywhere with a unified experience,” it also claims in that statement. As noted by the Harvard Business Review in 2022, SaaS companies are thought of as money-printing machines because they're on the monthly subscription model, like Netflix, but boring. A judge wasn't impressed with her AI apology. It's an interesting case study in AI agents and that whole "agency" thing... Google is sued over the AI podcast generation in NotebookLM.

Amazon Web Services has launched two credit programs worth up to $100 million to help federal agencies leverage AWS cloud services and generative AI technologies for applications ranging from battle management to quantum computing. “We are excited to pursue multiple pathways and initiatives that invest in the technologies and solutions that directly address Department of War's most pressing, real-world challenges,” David Fitzgerald, deputy undersecretary of the Army, said today in an AWS blog post. The AWS Genesis Accelerator Initiative will support scientific research by the U.S. Department of Energy, including the National Nuclear Security Administration, associated national laboratories, federal research organizations and private-sector organizations. Each initiative will provide up to $50 million in credits over the next three years for access to AWS cloud technology, training and technical expertise. Further details are available via AWS' web portal for government accelerator initiatives. Have a scoop that you'd like GeekWire to cover? AWS CEO Matt Garman on the value of in-person work; investing in startups; and the ROI from AI The AWS outage is a warning about the risks of digital dependance and AI infrastructure Generative AI tops cybersecurity in 2025 tech budget priorities, new AWS study finds Catch every headline in your inbox

Binti, a San Francisco-based startup that develops software tools for child welfare agencies, opened a new office on Seattle's Lake Union. “Opening our first-ever satellite office is an exciting next step for us — and the fact that it's right on the water doesn't hurt,” Binti co-founder and CEO Felicia Curcuru wrote in a video post on LinkedIn. Founded in 2016, Binti has approximately 85 employees and plans to hire about 30 more people this year. The talk in the Seattle region can sometimes focus on how founders and startups are leaving for Silicon Valley to join an ecosystem that is especially rich in AI talent and companies. Binti is opening its Seattle office just a stone's throw from the Fremont neighborhood — home to Google, Adobe, Salesforce/Tableau, Brinc Drones, PATH and others — and up the road from South Lake Union — home to Amazon, Meta, Apple and more Google offices. The startup has raised more than $60 million from investors including Founders Fund, First Round Capital, and Michael Dell. Have a scoop that you'd like GeekWire to cover? Tech Moves: Smartsheet CTO departs; Overland AI taps legal lead; Zillow, Comcast promote execs Photos: Inside the Allen Institute for AI's new HQ in Seattle's first mass-timber office building Industrious to open new co-working space in Seattle near tech-heavy South Lake Union Google sets long-term plan to exit Seattle's Fremont neighborhood, consolidate in South Lake Union