The AI Essays Are Moving Markets

(upbeat music) Welcome to the TechBrew Ride Home for Tuesday, February 24th, 2026. I'm Brian McCulloch today, that AIS AI shared with you yesterday sure got Wall Street's attention. Anthropics says Chinese models are training off of Claude, a significant new breakthrough in chip production technology. And as fun as that trifold phone might be, you probably wanna wait for later iterations of the form factor. Here's what you missed today in the world of tech. (upbeat music) Remember that SAI shared with you just yesterday from trainee research? There's a reason I've been sharing AISAs that get chatter of late. They tend to move markets all the sudden, quoting the journal, "The weeks long sell off in software stocks, "deepened Monday amid general unease "about the threat posed by AI. "Sophere makers apploven, crowd strike, "data dog and expedia were all among "the S&P 500's worst 10 performers on Monday "dropping at least 7%. "A state street fund tracking S&P's software "and services index fell more than 5%. "Several of Monday's biggest decliner's, "we're discussed in a now viral Sunday night post "from Sattrini research, which outlined "a hypothetical scenario where AI profoundly impacts "the economy in the near future." End quote. End quote, "insured with news. "Mastercard, American Express, Visa, "Synchrony, financial and capital one "are stocks that are all suffering "from severe selling pressure on Monday, "which were flagged in the scenario analysis "as facing headwinds from agents "looking to avoid fees and white color workers "being displaced." End quote. End quote, "The FT, Aries, Apollo, and Blackstone "all fell at least 5% while KKR lost 9% "as shares in the company's extended "a poor start to the year amid concerns "that market volatility caused by worries "about AI disruption could slow fundraising "and delay asset sales. "Blue Owl slid 3% bringing its fall for the year "to about 30%. "The software sector accounted for roughly 18% "of US private equity deal value in 2025 "according to pitch book data." But wait, there's more, not related to that essay, specifically, but definitely AI related, quoting CNBC. International Business Machine Stock is getting slammed to Monday, becoming the latest perceived victim of a rapidly developing AI technology after Anthropic said its CLAWD code tool could be used to modernize legacy systems that run Cobra. Shares of IBM closed the day lower by nearly 13.2%. After Anthropic on Monday said CLAWD code could be used to automate the exploration and analysis work that drives most of the complexity and CobraWD modernization a key IBM business. IBM has long sold mainframe systems that are optimized for large-scale transaction processing where CobraWD has often been used. Short for common business-oriented language, CobraWD is a dominant code system developed in the late 1950s, often used in business data processing, such as payment processing and retail transaction systems. An estimated 95% of ATM transactions in the US use CobraWD, according to Anthropic, making it a prime target for cost-efficient AI disruption. Hundreds of billions of lines of CobraWD run in production every day, powering critical systems and finance airlines and government. Despite that, the number of people who understand it shrinks every year Anthropic wrote in a Monday blog post, AI excels at streamlining the tasks that once made CobraWD modernization cost prohibitive. CLAWD code can help modernize CobraWD code bases by mapping dependencies across thousands of lines of code, documenting workflows and identifying risks that would take human analysts months to surface Anthropic said. Legacy code modernization stalled for years because understanding legacy code costs more than rewriting it. AI flips that equation the blog post said, end quote. (upbeat music) Anthropic says deep seek, minimax and moonshot violated its terms of service by prompting CLAWD a combined more than 16 million times and using distillation to train their own products. Quoting the journal. Earlier this month, the Anthropic rival OpenAI sent a memo to house lawmakers accusing deep seek of using the same tactic called distillation to mimic OpenAI's products. Anthropic said distillation had legitimate uses, companies use it to build smaller versions of their own products, for example, but it could also be used to build competitive products in a fraction of the time and at a fraction of the cost. The scale of the different companies distillation activity varied, deep seek engaged in 150,000 interactions with CLAWD whereas moonshot and minimax had more than 3.4 million and 13 million respectively Anthropic said. Representatives from deep seek moonshot and minimax didn't respond to requests for comment. Many Chinese companies including moonshot and minimax have recently released their latest AI models, many of which feature enhanced reasoning and coding capabilities. DeepSeek is preparing to roll out its next generation model soon. In a research paper updated in September, deep seek said that during a late stage of pre-training its flagship V3 model, it exclusively used plain web pages and e-books without incorporating any synthetic data. However, it said some web pages contained a significant number of OpenAI model generated answers. DeepSeek said its base model might have acquired knowledge from other powerful models indirectly by drawing on such web pages. Synthetic data often using distillation has been increasingly adopted for training large foundation models as developers face a shortage of high quality data and focus on giving models so-called agentic capabilities, meaning allowing them to take action proactively to complete tasks on behalf of users. In a technical report in July, moonshot said it used synthetic data for training its Kimi K2 model. Anthropic said the activity by the Chinese developers raised national security concerns for the US, foreign labs that distill American models can then feed these unprotected capabilities into military, intelligence, and surveillance systems, the company said, and quote. [MUSIC PLAYING] Meta has agreed to acquire up to six gigawatts of AMD instinct GPUs and a deal valued at more than $100 billion that could see Meta own up to 10% of AMD the company. Meta plans to deploy one gigawatts of this by the end of this year. Quoting the journal shares an AMD Rose 14% in pre-market trading under Tuesday's agreement, Meta will buy enough of AMD's latest chips known as the MI450 series to power data centers using up to six gigawatts of computing power over the next five years. Each gigawatt of computing power means several tens of billions of dollars in revenue for AMD, the company said. Meta is expected to deploy the first gigawatts starting later this year. As part of the arrangement, AMD has agreed to give Meta warrants to buy up to 160 million AMD shares or roughly 10% of the company for a penny a piece as long as certain milestones are met. The full stock award is conditional on a rise in AMD's share price. Meta would only receive the final trunch of shares once AMD's stock hits $600. It closed at $196.60 Monday. With a limited number of large buyers of their chips, both Nvidia and AMD have been using novel financing mechanisms to lock key customers into long-term agreements to use their technology. In October, AMD struck a deal with OpenAI that had terms almost identical to those of the Meta partnership. Both deals are examples of what critics have dubbed circular financing and arrangement under which one company pays another company, which turns around and buys products or services from the first company. As large customers such as Meta increase spending on AI infrastructure, AMD is seeking to lock them into using its chips for as long as possible. Meta has a lot of choices. This CEO of AMD said, I want to make sure that we are always a clear seat at the table when they think about what they need next. Last week, Meta said it would buy several million of Nvidia's GPUs as well, that deal, which is expected to cost tens of billions of dollars is meant to accelerate the Facebook and Instagram owners met a compute effort under which the company is planning to increase capital spending on AI computing to develop large language models and further optimize its ad business. Meta plans to deploy tens of gigawatts of data center computing power this decade and hundreds of gigawatts or more over time. Chief Executive Mark Zuckerberg wrote in a social media post in January, the company spent $72 billion last year to build out AI data centers and plans to spend as much as $135 billion this year, end quote. ASML researchers have unveiled a breakthrough in EUV light source power, increasing output from 600 to 1000 watts, a jump that could yield 50% more chips by 2030. Quoting Reuters, ASML is the world's only maker of commercial extreme ultraviolet lithography EUV machines, a critical tool for chip makers such as Taiwan semiconductor manufacturing, Intel, and others in producing advanced computing chips. It's not a parlor trick or something like this, where we demonstrate for a very short time that it can work. Michael Purvis, ASML's lead technologist for its EUV source light said in an interview, it's a system that can produce 1000 watts under all the same requirements that you could see at a customer, he added, speaking at the company's California facilities near San Diego. The EUV machines are so critical to chip production that US governments of both parties have worked with Dutch leaders to prevent them from being shipped to China, spurring it to launch a national effort to build machines of its own. In the United States, at least two startups substrate and X-light have raised hundreds of millions of dollars to develop American competitors to ASML's technology with X-light securing government funding from President Donald Trump's administration. With the technological advance revealed on Monday, which is being reported here for the first time, ASML aims to out-distance any would-be rivals by improving the most technologically challenging aspects of the machines. This is the quest to generate EUV light with the right power and properties to turn out chips at high volume. The company's researchers have found a way to boost the power of the EUV light source to 1000 watts from 600 watts now. The chief advantage is that greater power translates into the ability to make more chips every hour helping to lower the cost of each. Chips are printed similar to a photograph where the EUV light is shown on a silicon wafer coated with special chemicals called a photoresist. With more powerful EUV light source chip factories need shorter exposure times. We'd like to make sure that our customers can keep on using EUV at a much lower cost. Ton Van Gogh, executive vice president for the NXC line of EUV machines at ASML told Reuters, ASML got the power boost by doubling down on an approach that already places its machines among the most complex inventions of humans. To produce light with a wavelength of 13.5 nanometers, ASML's machine shoots a stream of molten droplets of tin through a chamber where a massive carbon dioxide laser heats them into plasma. This is a superheated state of matter in which the tin droplets become hotter than the sun and emit EUV light to be collected by precision optic equipment supplied by Germany's Carl Zeiss AG and fed into the machine to print chips. The key advancements in Monday's disclosure involved doubling the number of tin droplets to about 100,000 every second and shaping them into plasma using two smaller laser bursts as opposed to today's machines that use a single shaping burst and quote. Finally today, the first review of Samsung's $2,899 Galaxy Z Trifold. Chris Welch at Bloomberg says the trifold works well as a smartphone, especially for entertainment, but falls short as a tablet and laptop replacement by some unpredictability quote. Like all first generation products, the trifold is essentially a love letter to gadget nerds like me. This is not a smartphone for consumers who are rational about their tech purchases. No one other than well-healed early adopters should give thought to buying one. I've used the Galaxy Z Fold 7 constantly since its release last summer, that device's lightweight thin design finally overcame the trade-offs of previous book style foldables, but one weakness is watching video since the aspect ratio is nearly square. When it's open, there are thick black bars above and below most movies and TV shows you watch. Content is slightly bigger than on an iPhone 17 Pro Max, for example, but just barely. The trifold solves that problem and delivers an entertainment experience that outclasses any other phone on the market. A wider screen means letterboxing is far less pronounced and video appears substantially bigger. For those who can't get enough TikTok or Instagram or even if you're not using the trifold, the difference between the Z Fold 7's 8-inch display and the trifold's 10-inch inner panel doesn't sound like much, but it's noticeable in practice. The trifold is heavier and thicker than a normal smartphone when closed shut and you definitely feel it. It can be unwieldy for those with small hands, but when open, that same weight seems light compared with conventional 10-inch tablets. The trifold is undeniably in the way of a smartphone. The trifold is undeniably an impressive piece of technology and an even better entertainment consumption device. There's no other smartphone available in the US that can transform into an immersive 10-inch wide screen display in a matter of seconds. The problem is, I rarely need both of those things at the same exact moment. My 13-inch M5 iPad Pro is more powerful and provides extra room to work with. Its display is brighter and punchier than the trifold screen and creaseless. I can't even grab it into my jeans pocket, but I'm fine just grabbing the tablet when I need it. Even if you can comfortably afford the trifold, it's best to hold off. Apple's forthcoming foldable iPhone is believed to have a squat form factor that would match the trifold's appeal as a portable video gadget. Samsung itself is rumored to be developing a wide version of its upcoming Galaxy Z Fold 8, a better video-watching experience is already on the horizon and it will cost far less than $2,899. The pocket computer, the relatively cheaper $2,000 Galaxy Z Fold 7, already offers similar multitasking in a much lighter package. Plus its inner screen is still ideal for reading, writing, in Google Docs, browsing the web, or touching up photos. In both of these cases, you won't get an experience quite as large or as immersive as what a dual-inch trifold can deliver, but I can't justify spending so much on a smartphone when Android itself has some catching up to do. As a lover of bleeding-edge tech, part of me will be sad to return my trifold review unit now that my testing is done. I've enjoyed using it, even with the creases, average camera, and other hardware quirks like the phone's constant table wobble. But if I'm ever going to pay this much for a phone, it should have barely any compromises or downsides. Like any first-generation product, the trifold has its share of them. It's unclear how frequently we'll see devices like this. The release cadence is certain to be slower than regular foldables and there's no indication that Google and Motorola to keep players in that space will bother with trifolds. So it will be up to Samsung and Chinese phone makers to show whether this form factor has staying power or is simply an engineering flex. For the time being, laptop and tablet makers don't have to worry about being supplanted and ". Nothing more for you today. Talk to you tomorrow."