Anthropic's fourth installment of its Economic Index, drawing on an anonymized sample of two million Claude conversations from November 2025, finds that AI is changing how people work rather than whether they work at all. The study tracked usage across the company's consumer-facing Claude.ai platform and its API, categorizing interactions as either automation (where AI completes tasks entirely) or augmentation (where humans and AI collaborate). The split came out to 52% augmentation and 45% automation on Claude.ai, a slight shift from January 2025 when augmentation led 55% to 41%.

The share of jobs using AI for at least a quarter of their tasks has risen from 36% in January to 49% across pooled data from multiple reports. Anthropic's researchers also found that AI delivers its largest productivity gains on complex work requiring college-level education, speeding up those tasks by a factor of 12 compared to 9 for high-school-level work.

Claude completes college-degree tasks successfully 66% of the time versus 70% for simpler work. Computer and mathematical tasks continue to dominate usage, accounting for roughly a third of Claude.ai conversations and nearly half of API traffic.

White-collar workers stuck in a cycle of layoffs and stagnant wages might want to look past the traditional tech, finance and media job postings to an unexpected source of opportunity: the blue-collar sector, which faces a labor shortage and is seeing rapid transformation through private-equity investment. These jobs are generally less vulnerable to AI, and the earning trajectory can be steep, the WSJ writes.

At Crash Champions, a car-repair chain that has grown from 13 locations in 2019 to about 650 shops across 38 states, service advisers start at roughly $60,000 after a six-month apprenticeship and can double that within 18 months, according to CEO Matt Ebert. Directors overseeing multiple locations earn more than $200,000. Power Home Remodeling, a PE-backed construction company, says tech sales professionals earning $85,000 to $100,000 could make lateral moves after a 10-week training program.

The share of workers in their early 20s employed in blue-collar roles rose from 16.3% in 2019 to 18.4% in 2024, according to ADP -- five times the increase among 35- to 39-year-olds.

An anonymous reader quotes a report from TechCrunch: Over the weekend, Neel Somani, who is a software engineer, former quant researcher, and a startup founder, was testing the math skills of OpenAI's new model when he made an unexpected discovery. After pasting the problem into ChatGPT and letting it think for 15 minutes, he came back to a full solution. He evaluated the proof and formalized it with a tool called Harmonic -- but it all checked out. "I was curious to establish a baseline for when LLMs are effectively able to solve open math problems compared to where they struggle," Somani said. The surprise was that, using the latest model, the frontier started to push forward a bit.

ChatGPT's chain of thought is even more impressive, rattling off mathematical axioms like Legendre's formula, Bertrand's postulate, and the Star of David theorum. Eventually, the model found a Math Overflow post from 2013, where Harvard mathematician Noam Elkies had given an elegant solution to a similar problem. But ChatGPT's final proof differed from Elkies' work in important ways, and gave a more complete solution to a version of the problem posed by legendary mathematician Paul Erdos, whose vast collection of unsolved problems has become a proving ground for AI.

For anyone skeptical of machine intelligence, it's a surprising result -- and it's not the only one. AI tools have become ubiquitous in mathematics, from formalization-oriented LLMs like Harmonic's Aristotle to literature review tools like OpenAI's deep research. But since the release of GPT 5.2 -- which Somani describes as "anecdotally more skilled at mathematical reasoning than previous iterations" -- the sheer volume of solved problems has become difficult to ignore, raising new questions about large language models' ability to push the frontiers of human knowledge. Somani examined the online archive of more than 1,000 Erdos conjectures. Since Christmas, 15 Erdos problems have shifted from "open" to "solved," with 11 solutions explicitly crediting AI involvement.

On GitHub, mathematician Terence Tao identifies eight Erdos problems where AI made meaningful autonomous progress and six more where it advanced work by finding and extending prior research, noting on Mastodon that AI's scalability makes it well suited to tackling the long tail of obscure, often straightforward Erdos problems.

Progress is also being accelerated by a push toward formalization, supported by tools like the open-source "proof assistant" Lean and newer AI systems such as Harmonic's Aristotle.

Games Workshop, the owner and operator of a number of hugely popular tabletop war games, including Warhammer 40,000 and Age of Sigmar, has banned the use of generative AI in its content and design processes. IGN reports: Delivering the UK company's impressive financial results, CEO Kevin Rountree addressed the issue of AI and how Games Workshop is handling it. He said GW staff are barred from using it to actually produce anything, but admitted a "few" senior managers are experimenting with it. Rountree said AI was "a very broad topic and to be honest I'm not an expert on it," then went on to lay down the company line:

"We do have a few senior managers that are [experts on AI]: none are that excited about it yet. We have agreed an internal policy to guide us all, which is currently very cautious e.g. we do not allow AI generated content or AI to be used in our design processes or its unauthorized use outside of GW including in any of our competitions. We also have to monitor and protect ourselves from a data compliance, security and governance perspective, the AI or machine learning engines seem to be automatically included on our phones or laptops whether we like it or not.

We are allowing those few senior managers to continue to be inquisitive about the technology. We have also agreed we will be maintaining a strong commitment to protect our intellectual property and respect our human creators. In the period reported, we continued to invest in our Warhammer Studio -- hiring more creatives in multiple disciplines from concepting and art to writing and sculpting. Talented and passionate individuals that make Warhammer the rich, evocative IP that our hobbyists and we all love."

Digg is officially back under the ownership of its original founder, Kevin Rose, along with Reddit co-founder Alexis Ohanian. "Similar to Reddit, the new Digg offers a website and mobile app where you can browse feeds featuring posts from across a selection of its communities and join other communities that align with your interests," reports TechCrunch. "There, you can post, comment, and upvote (or 'digg') the site's content." From the report: [T]he rise of AI has presented an opportunity to rebuild Digg, Rose and Ohanian believe, leading them to acquire Digg last March through a leveraged buyout by True Ventures, Ohanian's firm Seven Seven Six, Rose and Ohanian themselves, and the venture firm S32. The company has not disclosed its funding. They're betting that AI can help to address some of the messiness and toxicity of today's social media landscape. At the same time, social platforms will need a new set of tools to ensure they're not taken over by AI bots posing as people.

"We obviously don't want to force everyone down some kind of crazy KYC process," said Rose in an interview with TechCrunch, referring to the 'know your customer' verification process used by financial institutions to confirm someone's identity. Instead of simply offering verification checkmarks to designate trust, Digg will try out new technologies, like using zero-knowledge proofs (cryptographic methods that verify information without revealing the underlying data) to verify the people using its platform. It could also do other things, like require that people who join a product-focused community verify they actually own or use the product being discussed there.

As an example, a community for Oura ring owners could verify that everyone who posts has proven they own one of the smart rings. Plus, Rose suggests Digg could use signals acquired from mobile devices to help verify members -- for instance, the app could identify when Digg users attended a meetup in the same location. "I don't think there's going to be any one silver bullet here," said Rose. "It's just going to be us saying ... here's a platter of things that you can add together to create trust."

Cerebras Systems landed a more than $10 billion deal to supply up to 750 megawatts of compute to OpenAI through 2028, according to a blog post by OpenAI. CNBC reports: The deal will help diversify Cerebras away from the United Arab Emirates' G42, which accounted for 87% of revenue in the first half of 2024. "The way you have three very large customers is start with one very large customer, and you keep them happy, and then you win the second one," Cerebras' co-founder and CEO Andrew Feldman told CNBC in an interview.

Cerebras has built a large processor that can train and run generative artificial intelligence models. [...] "Cerebras adds a dedicated low-latency inference solution to our platform," Sachin Katti, who works on compute infrastructure at OpenAI, wrote in the blog. "That means faster responses, more natural interactions, and a stronger foundation to scale real-time AI to many more people."

The deal comes months after OpenAI worked with Cerebras to ensure that its gpt-oss open-weight models would work smoothly on Cerebras silicon, alongside chips from Nvidia and Advanced Micro Devices. OpenAI's gpt-oss collaboration led to technical conversations with Cerebras, and the two companies signed a term sheet just before Thanksgiving, Feldman said in an interview with CNBC. The report notes that this deal helps strengthen Cerebras' IPO prospects. The $10+ billion OpenAI deal materially improves revenue visibility, customer diversification, and strategic credibility, addressing key concerns from its withdrawn filing and setting the stage for a more compelling refile with updated financials and narrative.

The U.S. has approved limited sales of Nvidia's H200 AI chips to China, the Department of Commerce said on Tuesday. Exports will be allowed to "approved customers" with security safeguards and a 25% U.S. government cut. The company's most advanced Blackwell chips will remain restricted. The BBC reports: The H200, Nvidia's second-most-advanced semiconductor, had been restricted by Washington over concerns that it would give China's technology industry and military an edge over the U.S. The Commerce Department said the chips can be shipped to China granted that there is sufficient supply of the processors in the U.S.

Nvidia's spokesperson told the BBC that the company welcomed the move, saying it will benefit manufacturing and jobs in the U.S. The Commerce Department's Bureau of Industry and Security said its revised export policy applies to Nvidia's H200 chips, as well as less advanced processors. Chinese customers must also show "sufficient security procedures" and cannot use the chips for military uses.

Chinese embassy spokesman Liu Pengyu told the BBC on Wednesday that Beijing has consistently opposed the "politicization and weaponization of tech and trade issues." "We oppose blocking and restricting China, which disrupts the stability of industrial and supply chains," he said. "This approach does not serve the common interests of both sides."

Bandcamp has announced a ban on music made wholly or substantially by generative AI, aiming to protect human creativity and prohibit AI impersonation of artists. Here's what the music platform had to say: ... Something that always strikes us as we put together a roundup like this is the sheer quantity of human creativity and passion that artists express on Bandcamp every single day. The fact that Bandcamp is home to such a vibrant community of real people making incredible music is something we want to protect and maintain. Today, in line with that goal, we're articulating our policy on generative AI. We want musicians to keep making music, and for fans to have confidence that the music they find on Bandcamp was created by humans.

Our guidelines for generative AI in music and audio are as follows:
- Music and audio that is generated wholly or in substantial part by AI is not permitted on Bandcamp.
- Any use of AI tools to impersonate other artists or styles is strictly prohibited in accordance with our existing policies prohibiting impersonation and intellectual property infringement.

If you encounter music or audio that appears to be made entirely or with heavy reliance on generative AI, please use our reporting tools to flag the content for review by our team. We reserve the right to remove any music on suspicion of being AI generated. We will be sure to communicate any updates to the policy as the rapidly changing generative AI space develops. Given the response around this to our previous posts, we hope this news is welcomed. We wish you all an amazing 2026. [...]

Dell's chief operating officer Jeff Clarke has informed employees that the company is preparing for what he calls the "biggest transformation in company history," a sweeping systems overhaul scheduled to launch on May 3 that will standardize processes across nearly every major division.

The initiative, dubbed One Dell Way, will replace Dell's existing sprawl of applications, servers and databases with a single enterprise platform designed to unify the 42-year-old company's operations. Clarke's memo, sent to staff on Tuesday and obtained by Business Insider, said Dell has spent the past two years building toward this transition.

The May 3 launch will affect the company's PC business, finance, supply chain, marketing, sales, revenue operations, services, and HR. The ISG division, which handles cloud and AI infrastructure, will follow in August. "We need one way -- simplified, standardized and automated -- so we can be more competitive and serve our customers better," Clarke wrote. Mandatory training begins February 3.

McKinsey is asking graduate applicants to "collaborate" with an AI tool as part of its recruitment process, as competence with the technology becomes a requirement in competing for top-level jobs. From a report: The blue-chip consultancy is incorporating an "AI interview" into some final-round interviews, according to CaseBasix, a US company that helps candidates apply for posts at leading strategic consulting companies.

In an online post, CaseBasix said candidates in "select final rounds" in the US have been asked to complete tests using McKinsey's internal AI tool, Lilli. They are required to carry out practical consulting tasks with the help of Lilli. "In the McKinsey AI interview, you are expected to prompt the AI, review its output, and apply judgment to produce a clear and structured response. The focus is on collaboration and reasoning rather than technical AI expertise," CaseBasix said.

Amazon founder Jeff Bezos once told an audience that he views local PC hardware the same way he views a 100-year-old electric generator he saw in a brewery museum -- as a relic of a pre-grid era, destined to be replaced by centralized utilities that users simply rent rather than own. The anecdote, shared at a talk a few years ago, positioned Amazon Web Services and Microsoft Azure as the inevitable successors to the desktop tower. Bezos argued that users would eventually abandon local computing for cloud-based solutions, much as businesses once abandoned on-site power generation for the electrical grid.

Current market dynamics have made that prediction feel more plausible. DRAM prices have become increasingly untenable for consumers, and companies like Dell and ASUS have signaled price increases across their PC ranges. Micron has shut down its consumer DRAM operations entirely, prioritizing AI datacenter demand instead. SSD storage is expected to face similar constraints. Cloud gaming services from Amazon Luna, NVIDIA GeForce Now and Xbox are seeing steady growth.

Microsoft previously developed a consumer version of its business-grade Windows 365 cloud PC product, though the company deprioritized it -- the economics didn't work when cheap laptops remained available. That calculus could shift. Xbox Game Pass's 1440p cloud gaming runs $30 monthly and NVIDIA recently imposed a 100-hour cap on its cloud platform. The infrastructure remains expensive to operate, but rising local hardware costs may eventually close that gap.

Matthew McConaughey is taking a novel legal approach to combat unauthorized AI fakes: trademarking himself. From a report: Over the past several months, the "Interstellar" and "Magic Mike" star has had eight trademark applications approved by the U.S. Patent and Trademark Office featuring him staring, smiling and talking. His attorneys said the trademarks are meant to stop AI apps or users from simulating McConaughey's voice or likeness without permission -- an increasingly common concern of performers.

The trademarks include a seven-second clip of the Oscar-winner standing on a porch, a three-second clip of him sitting in front of a Christmas tree, and audio of him saying "Alright, alright, alright," his famous line from the 1993 movie "Dazed and Confused," according to the approved applications. "My team and I want to know that when my voice or likeness is ever used, it's because I approved and signed off on it," the actor said in an email. "We want to create a clear perimeter around ownership with consent and attribution the norm in an AI world."

The chief constable of one of Britain's largest police forces has admitted that Microsoft's Copilot AI assistant made a mistake in a football (soccer) intelligence report. From a report: The report, which led to Israeli football fans being banned from a match last year, included a nonexistent match between West Ham and Maccabi Tel Aviv.

Copilot hallucinated the game and West Midlands Police included the error in its intelligence report without fact checking it. "On Friday afternoon I became aware that the erroneous result concerning the West Ham v Maccabi Tel Aviv match arose as result of a use of Microsoft Co Pilot [sic]," says Craig Guildford, chief constable of West Midlands Police, in a letter to the Home Affairs Committee earlier this week. Guildford previously denied in December that the West Midlands Police had used AI to prepare the report, blaming "social media scraping" for the error.

Meta is shutting down three acquired VR studios as part of Reality Labs layoffs and a strategic pivot away from VR content toward AI-powered smart glasses. UploadVR reports: Meta shut down Twisted Pixel Games (Deadpool VR), Sanzaru Games (Asgard's Wrath), and Armature Studio (Resident Evil 4 VR). [...] Twisted Pixel Games was founded in 2006 and mostly made Xbox games published by Microsoft for the first decade of its existence. In fact, Microsoft owned the studio from 2011 until 2015, when it became an independent company again. On contract from Facebook, between 2017 and 2019 Twisted Pixel released four VR games: Wilson's Hearth (Rift); B-Team (Go/Quest); Defector (Rift); and Path of the Warrior (Rift/Quest). In 2022, Twisted Pixel Games was acquired by Meta. And just two months ago, it released what it had been working on since then: Deadpool VR, the latest Quest-exclusive VR game. [...]

Sanzaru Games was also founded in 2006, and made a combination of its own games and contract titles for companies such as Sony, porting the original God of War series to PS Vita. Sanzaru Games was also contracted by Facebook to build VR games for the Oculus Rift and its Touch controllers, between 2016 and 2019: Ripcoil (2016); VR Sports Challenge (2016); Marvel Powers United VR (2018); and Asgard's Wrath (2019). In 2020, Sanzaru Games was acquired by Facebook, and in 2023 released Asgard's Wrath 2, taking the core essence of Asgard's Wrath to Quest 2 and Quest 3 standalone, with a semi-open world and a campaign more than 60 hours long. Exactly one year ago, Sanzaru released the last major content update for Asgard's Wrath 2, stating that it was now working on the "next big thing" with no detail released on what that would be before the studio closed.

Founded in 2008, Armature Studio was mainly a porting studio, bringing PC titles to consoles and console titles to PS Vita. Like Twisted Pixel and Sanzaru, Armature too was contracted by Facebook to build early consumer VR games: Fail Factory (2017); Sports Scramble (2019); and Resident Evil 4 VR (2021). Armature was acquired by Meta in 2022, and many VR gamers had been eagerly anticipating what it had been working on since. Whatever it was, Armature too is now shut down.

The U.S. Senate unanimously passed the Disrupt Explicit Forged Images and Non-Consensual Edits Act (DEFIANCE Act), giving victims of sexually explicit AI deepfakes the right to sue the individuals who created them. The Verge reports: The bill passed with unanimous consent -- meaning there was no roll-call vote, and no Senator objected to its passage on the floor Tuesday. It's meant to build on the work of the Take It Down Act, a law that criminalizes the distribution of nonconsensual intimate images (NCII) and requires social media platforms to promptly remove them. [...] Now the ball is again in the House leadership's court; if they decide to bring the bill to the floor, it will have to pass in order to reach the president's desk.

An anonymous reader quotes a report from Network World: The global technology sector eliminated some 244,851 jobs in 2025, according to a report from RationalFX. The U.K.-based financial services company says the worldwide downsizing reflects how companies in 2025 restructured their operations to focus on efficiency, profitability, and AI-driven productivity. The RationalFX analysis, which examined layoffs reported by TrueUp, TechCrunch, and multiple state WARN databases, points to economic uncertainty, elevated interest rates, and accelerating AI and automation adoption as reasons that 2025 marked "another year of sustained downsizing following the post-pandemic correction that began in 2022."

Companies indicated that AI and automation were among the most frequently cited drivers for layoffs in 2025. Some companies retrained employees when faced with the technology; many replaced roles entirely, RationalFX reports. "Tech sector layoffs in 2025 displaced hundreds of thousands of workers worldwide as companies accelerated structural resets rather than short-term cost corrections," said Alan Cohen, analyst at RationalFX, in a statement. "While macroeconomic pressures such as high interest rates, trade restrictions, and geopolitical uncertainty continued to weigh on business confidence, the dominant force behind last year's job cuts was the rapid adoption of automation and artificial intelligence."

The analysis also uncovered that U.S.-headquartered technology companies were responsible for the majority of job losses, accounting for approximately 69.7% of all global tech layoffs. This resulted in more than 170,000 employees being cut across both domestic and offshore operations from U.S. tech companies. California spearheaded layoffs in the U.S. tech sector this year, with 73,499 job cuts accounting for roughly 43.08% of all tech layoffs in the country, according to the RationalFX report. The report also points out that Washington has seen 42,221 tech jobs cut since the start of the year, accounting for 24.74% of all U.S. tech layoffs. Intel contributed the single largest number of layoffs last year, reducing its headcount from 109,000 people at the end of 2024 to around 75,000 by the end of 2025. Other major U.S. tech companies with large-scale layoffs last year include Amazon (more than 20,000 jobs cut), Microsoft (approximately 19,215 layoffs), Verizon (15,000 employees), Accenture (11,000 employees), IBM (9,000 job cuts), and HP (6,000 roles).

An anonymous reader quotes a report from MacRumors: Apple today introduced a new Apple Creator Studio bundle that offers access to six creative apps, as well as exclusive AI features and content, as part of a single subscription. In the U.S., pricing is set at $12.99 per month or $129 per year. Here are the six apps included with an Apple Creator Studio subscription: Final Cut Pro on the Mac and iPad; Logic Pro on the Mac and iPad; Pixelmator Pro on the Mac and iPad; Motion on the Mac; Compressor on the Mac; and MainStage on the Mac.

Pixelmator Pro was previously only available on the Mac, but it is coming to the iPad. Apple Creator Studio subscribers will also receive access to exclusive AI features and premium content across not only the Final Cut Pro and Pixelmator Pro apps, but also the iWork apps Numbers, Pages, and Keynote, and the Freeform app later this year. So if you want the best, fully-featured versions of all of these apps going forward, you will need to subscribe to the bundle.

Apple says there will be separate Creator Studio and one-time purchase "versions" of each app. If you have both versions installed on your Mac, the Creator Studio versions will have "unique icons" so that they stand out, according to Apple. Apple Creator Studio will be available through the App Store starting Wednesday, January 28. All new subscribers will be able to receive a one-month free trial, and customers who purchase a new Mac or a qualifying iPad model with an A16, A17 Pro, or M-series chip or later will be eligible for an extended three-month free trial. "If you are not interested in subscribing to the new Apple Creator Studio bundle introduced today, you will officially start to miss out on some new features," adds MacRumors in a separate article. If you bought the apps via a one-time purchase, or plan to do so in the future, "you will no longer have access to all new features," though they will continue to receive updates.

"There are some exceptions, as Apple says Logic Pro and MainStage will have all the same features whether they are subscription or one-time-purchase versions," notes MacRumors. "It looks like most if not all of the new features that will be limited to Creator Studio subscribers will be powered by AI, as Apple repeatedly describes them as 'intelligent' features. The apps are continuing to receive other new features that do not require a subscription over time, so one-time purchasers are not completely left out."

America's largest power-grid operator, PJM, which delivers electricity to 67 million people across a 13-state region from New Jersey to Kentucky, is approaching a supply crisis as AI data centers in Northern Virginia's "Data Center Alley" consume electricity at an unprecedented rate.

The nonprofit expects demand to grow by 4.8% annually over the next decade. Mark Christie, former chairman of the Federal Energy Regulatory Commission, said the reliability risk that was once "on the horizon" is now "across the street." Dominion Energy, the utility serving parts of Virginia, has received requests from data-center developers requiring more than 40 gigawatts of electricity -- roughly twice its Virginia network capacity at the end of 2024. Older power plants are going out of service faster than new ones can be built, and the grid could max out during periods of high demand, forcing rolling blackouts during heat waves or deep freezes.

In November, efforts to establish new rules for data centers stalled when PJM, tech companies, power suppliers and utilities couldn't agree on a plan. Monitoring Analytics, the firm that oversees the market, warned that unless data centers bring their own power supply, "PJM will be in the position of allocating blackouts rather than ensuring reliability."

Moxie Marlinspike, the engineer who created Signal Messenger and set a new standard for private communications, is now trialing Confer, an open source AI assistant designed to make user data unreadable to platform operators, hackers, and law enforcement alike. Confer relies on two core technologies: passkeys that generate a 32-byte encryption keypair stored only on user devices, and trusted execution environments on servers that prevent even administrators from accessing data. The code is open source and cryptographically verifiable through remote attestation and transparency logs.

Marlinspike likens current AI interactions to confessing into a "data lake." A court order last May required OpenAI to preserve all ChatGPT user logs including deleted chats, and CEO Sam Altman has acknowledged that even psychotherapy sessions on the platform may not stay private.

Meta has begun laying off more than 1,000 employees from its Reality Labs division as the company redirects resources away from virtual reality and metaverse products toward AI wearables and smartphone features. The cuts amount to roughly 10% of Reality Labs' 15,000-person workforce, according to an internal post from CTO Andrew Bosworth reviewed by Bloomberg.

Reality Labs has lost more than $70 billion since the start of 2021, and top executives discussed budget cuts as deep as 30% for the metaverse group in December. Meta plans to continue developing its Horizon metaverse platform, but the focus will shift almost exclusively to mobile phones rather than the fully immersive VR headsets the company originally envisioned.