Everyone Said AI Would Kill Google. Its First-Ever $100 Billion Quarter Just Proved Them Wrong

Every story about Google starts and ends with search. It makes sense—Google is a search engine. It’s not only the company’s most important and profitable business, but it’s the thing that has defined the internet for two decades. But for most of the past two years, the biggest story about Google has been that artificial intelligence would, inevitably, make search obsolete. People would stop “Googling” things because AI chatbots could just tell them the answers. Search—the company’s $200-billion-a-year cash cow—was supposed to be doomed. On the one hand, the idea that people would no longer type queries into Google’s search box and then click on the blue links that show up on results pages was a doomsday scenario. And AI chatbots certainly made that look increasingly likely. Then again, that story always assumed Google would sit still while the world around it changed. It assumed the company that practically invented the modern internet—or at least the way most of us experience it—wouldn’t figure out how to adapt. On Wednesday, Alphabet, Google’s parent company, reported its first-ever $100 billion quarter. Revenue rose 16 percent to $102.3 billion. Net income jumped 33 percent to $34.98 billion. Those are not the numbers of a company whose main business is being disrupted. It’s more like the numbers of a company that’s quietly figuring out how to change with the behavior of its users. Google Search and YouTube each grew at a double-digit pace. “Google Search & other” revenue climbed 15 percent to $56.6 billion. YouTube ads rose 15 percent to $10.3 billion. Combined, Google’s advertising machine brought in more than $74 billion for the quarter. Not only that, but its cloud business grew by 35 percent over the previous year. That leads to the most interesting part of this story, which is the part about how Google is spending all that money. As it announced its earnings, Google said it would raise its capital expenditures, specifically as it invests in infrastructure to serve its cloud businesses. That’s the part of the business that powers its AI ambitions. Google made more money than ever from search, and it’s spending that money on AI. Training and running massive models requires staggering amounts of computing power. But that’s exactly where Google’s advantage lies—it already owns what is probably the largest global computing infrastructure ever built. Now, it’s doubling down. Alphabet expects to spend $91 billion to $93 billion in capital expenditures this year—mostly on data centers, networking, and custom chips designed for AI workloads. That’s up sharply from last year and puts Google in the same spending league as Amazon and Microsoft. And even with those huge investments, Alphabet’s operating margin—excluding a $3.5 billion European Commission fine—rose to 33.9 percent. In other words, it’s spending tens of billions to expand AI capacity while remaining one of the most profitable companies on the planet. Google’s strategy isn’t just about protecting search ads. It’s about using the strength of that business to fund a transformation into something bigger: the dominant AI platform. That’s still a big lift. Yes, Google is a household name, but it’s still behind in AI—at least in terms of consumer mindshare. OpenAI’s ChatGPT is the front-runner in terms of customer adoption, but Google has almost every other advantage. It has the technology, the infrastructure, and a built-in user base that already trusts it as the default source of information. And because Google controls so many layers of the stack—hardware, data centers, models, and consumer products—it can absorb the cost of AI adoption in a way startups and rivals can’t. It doesn’t have to rent the future on someone else’s platform; it’s already building it. Now, Google is doing something very few companies have ever pulled off: funding its own disruption without losing momentum. Search and YouTube are still massive profit engines, generating the cash Google needs to build the infrastructure for AI. Basically, Google doesn’t really care whether you type your queries into a search box or a chatbot window, as long as you keep asking it your questions. For all the hype about AI replacing search, this quarter makes one thing clear: Google’s biggest business isn’t dying. It’s evolving into something that could be far more lucrative. If the company’s $93 billion AI spending spree pays off the way Pichai expects, Google might have just figured out a better end of the story than search. EXPERT OPINION BY JASON ATEN, TECH COLUMNIST @JASONATEN

AI ‘Consulting’ Services Can Help Smaller Businesses, but Risks Persist

Consultancy firms can be very useful for growing businesses, giving new companies guidance or financial or management advice when needed, backed by experience and expertise. But for smaller enterprises with narrow margins, the cost of hiring top-rank consulting firms can be financially out of reach. Enter AI, according to a new report on Business Insider. In much the same way that generative AI tools promise to add, say, coding expertise to a small team or free up workers from mundane tasks to engage in more productive work, AI-powered “consultant” apps are emerging from a suite of Silicon Valley startups, with the goal of helping small firms carry out market research, analyze data, or smooth and optimize their business operations. Business Insider quotes Thomson Nguyen, co-founder and managing partner of Wyoming-based venture capital outfit Saga, on the phenomenon. These new AI consultancy players won’t be challenging big consulting firms any time soon, he thinks, simply because if you’re a “Fortune 500 company building AI infrastructure for your call center, you’ll still hire the Big Four,” because you’ll have the budget set aside and experience in working with third-party consultancies. So the real target for these startups is smaller companies, making under $100 million a year, which are too small to hire a McKinsey or Deloitte. Business Insider notes that AI apps like PromptQL, from Bangalore-based AI unicorn Hasura, are directly set up to tackle typical consultant roles—including analyzing a company’s internal data, and continually adapting over time. PromptQL even has a team of engineers that’ll help craft an AI analyst agent specifically to meet the needs of a client company. Hasura co-founder and CEO Tanmai Gopal admitted to Business Insider that it’s “not as good as a McKinsey consultant,” but it has the benefits of being “instant.” That’s the very opposite of the sometimes protracted process during which a consultant learns about their client company before tackling an analysis, since an AI can just be switched on and immediately wrestle with data. Among the top tasks being assigned to AI consultancy startups are starting and managing call centers and customer service automation, integrating software and AI into the client company’s operations, and building management and operational AI systems. There are even AI tools targeting executive coaching. This may not be a surprise, considering that big tech names like Salesforce are already selling their own agent-based AI services aimed at automating the sales process and call center operations. AI startups offering similar options and targeting smaller companies as clients is natural. Gopal told Business Insider that, for now, these AI consultancy tools aren’t really replacing human workers—echoing many an AI evangelist’s arguments about the role of AI in the workplace. Human workers have more diverse skills, and for the moment such services depend as much on the “network” of colleagues that a human worker can access as on their advice. What’s the takeaway for your company? If you find yourself struggling with an expertise gap, you may find that there’s an AI-powered consulting tool out there that will fit your needs. But as with most AI tools, perhaps the thing to remember is that (just as with human consultants, though perhaps less obviously) AIs are not infallible. AI systems regularly make mistakes, and can hallucinate analysis and advice that they then pass off just as if it was real advice. You’ll have seen this by now, perhaps when you asked an AI to write a snippet of code for you. The AI may insist the code works, but when you say “No, it doesn’t,” the AI may say “Oh! You’re right!” and offer a fix. Whenever you’re using AI it’s best to run the results past a human worker before making a critical decision based on an AI consultant’s analysis. BY KIT EATON @KITEATON

Far From Silicon Valley, This Founder’s Data Center Business Is Building the Future of AI

There’s a common refrain among business strategists that it’s better to be a pickaxe salesman than a gold prospector—or, in other words, that the best way to capitalize on a gold rush is to sell tools to the people hoping to strike it rich, rather than trying to hit paydirt yourself. With artificial intelligence currently enjoying a boom of its own—one that some big names have even called a bubble—plenty of companies are stepping in to sell the AI equivalent of pickaxes, such as backend infrastructure and computational power. That includes big-name ventures such as CoreWeave, the AI cloud company that IPO’d earlier this year and counts Microsoft, IBM and OpenAI among its clientele. CoreWeave claimed the No. 45 spot on this year’s Inc. 5000 list of the fastest-growing private companies in America, and is one of the AI boom’s biggest winners so far. But beating it out on that same list was another AI infrastructure company, the Chicago-based Introl, which came in at No. 14 on the list just a few years after its founding. Introl founder and CEO Ryan Puckett launched the company in 2021 while between jobs and looking to work as a freelance project manager. Introl helps set up GPUs, or graphics processing units: the computer chips that train and run modern AI models. A former low-voltage cable technician, Puckett has since built the company up to impressive scale: it’s grown annual revenue nearly 10,000 percent over the last three years, and Puckett says domestic revenue was about $38 million last year. All of that growth is bootstrapped, the success of which the CEO attributes to “managing cashflow effectively and efficiently”—as well as, at least initially, “a lot of credit card debt.” And though Introl was born in Dallas, Puckett moved it to Chicago a few years in; he’d lived in the Windy City during his early 20s, and wanted to go back. “There’s not a better city in the country,” he says of Chicago. “There was no other thought in my mind to build it anywhere else.” Blake Crosley, Introl’s CTO, says the company has deployed “up to 100,000 GPU units in a data center.” Each one needs multiple connections, he adds, requiring lots and lots of fiber optic cable; the company says it has run more than 40,000 miles of the stuff in all. “We don’t actually own or operate the data centers,” Crosley explains. “We basically help design, like, what does it look like to actually get that set up in the space? Once the racks are in place, how are we going to actually connect everything together?” This work, he adds, is known as “rack and stack.” Installation is followed by testing and quality control. NDAs limit Introl’s ability to disclose specific client names, but the company says it has around 45 to 50 full-time employees, plus over 1,000 subcontractors. The startup deploys that workforce to data centers around the country and the planet. Those data centers are so big, Puckett says, that people get around them in golf carts and measure their footprints in terms of how many Costcos could fit inside. Speed to market is the CEO’s biggest challenge, he tells Inc. Companies will sometimes give Introl barely a week’s notice to get people on-site to a data center, he says, and it can sometimes be hard to find enough hotel space to house all those staffers—who will sometimes number in the hundreds for a job—especially in the small towns where many data centers go up. “In a lot of cases, because they are trying to get things online so quickly…certain specific sections of [the data centers] are being built while you’re in a different part,” Puckett says. “It’s a constant flow of trucks coming in, dropping off pallets of cables.” AI is big business right now, but if the fervor starts to die down, demand for the underlying hardware could follow suit too. “I’m not 100 percent sure what our pivot would be [if], say, GPU deployments just kind of fell off the face of the earth,” Puckett says, although Introl’s focus could shift toward maintenance. Right now, he estimates, 70 percent of the company’s work surrounds new installations, while the other 30 percent has to do with maintaining pre-existing sites. For now, though, the company is feeling good about where things are headed. “Obviously there’s a lot of talk about [an] AI bubble and stuff like that,” says Crosley, the CTO. “The players are huge, and the money that’s flowing is even bigger. But from a user perspective, on the side of utilizing AI, I can only see things expanding faster in the total adoption and usage.” BY BRIAN CONTRERAS @_B_CONTRERAS_

Here’s How LinkedIn Co-Founder Reid Hoffman Says AI Needs to Be Regulated

Regulation can be good for technology, so long as it’s done thoughtfully, according to LinkedIn co-founder, investor, and AI-enthusiast Reid Hoffman. Speaking on the heels of a pitch event in San Francisco called Entrepreneurs First Demo Day, he compared AI regulation to seatbelts in vehicles. “Seatbelts are a good thing, relative to the fact that regulatory stuff can have a positive impact on society, technology evolution. Now doing it smart in the right way is important,” he tells Inc. “You don’t try to solve everything before you get on the road. You get on the road and then solve it as you go,” he adds. His voice joins a chorus of others from big names in tech speaking up about how much—or in the case of legendary investor Marc Andreessen and companies like Meta—how little regulation they support. Hoffman sits on the board of Entrepreneurs First, an international talent investment firm that hosts incubator-style programs and related annual pitch competitions. Those events are called Demo Days, and the most recent took place in San Francisco on Wednesday. Hoffman joined EF’s board after leading a significant round of investment in the company in 2017 through his capacity at venture capital firm Greylock Partners. Hoffman was not on the ground at Demo Day this year, but another big name in tech was: Anthropic co-founder Jack Clark was the keynote speaker in conversation with Entrepreneurs First CEO Alice Bentinck. Just a few days prior, Clark had made waves for commentary he gave at The Curve conference in Berkeley, California, and later published in essay form in his newsletter. He compared AI to a “mysterious creature” of humanity’s own creation. He said he was optimistic about its potential as well as appropriately afraid of it, especially if AI’s goals are not absolutely aligned with humanity’s. And finally, he ended by emphasizing the need for conversations with a broad swathe of society to help craft a “policy solution.” “There will surely be some crisis,” Clark notes in his blog. “We must be ready to meet that moment both with policy ideas, and with a pre-existing transparency regime which has been built by listening and responding to people.” In response to the post, U.S. AI and crypto czar David Sacks accused Anthropic of fearmongering. Hoffman’s take, which he wrote about in his recent book, is by no means anti-regulation, but does differ somewhat from Clark’s. “In the book that I published in January, Superagency, part of what I was arguing for within AI is iterative deployment and development,” he tells Inc. “We do the regulatory thing, but we do it in response to what we can actually see versus imagination of what [could] happen,” he adds. AI has never been more topical, especially among aspiring entrepreneurs. This week at Demo Day in San Francisco, founders from 20 different startups pitched more than 200 tech investors, among them big name firms like a16z, Khosla Ventures, Paladin Capital, Insight Partners and Engine Ventures, in hopes of landing as much as $7 million in seed funding. It represented the culmination of some six months of work the founders had put in during Entrepreneurs First’s incubator-style program. On the lips of most of those entrepreneurs was AI. “The majority of the companies that were pitching yesterday—85 to 90 percent—are all using AI in some way. Some of them are building novel AI models, others are creating wrappers or scaffolding around existing AI models,” says Bentinck. “If you look at what early stage investors want to put capital behind, they see this enormous opportunity in the new AI economy.” Originally founded in London, Entrepreneurs First started off as a nonprofit in 2011 before becoming the investment vehicle it is today, starting in 2015. The company expanded overseas to offer programming in San Francisco at the start of 2024, and continues to run cohorts across Europe, India and the U.S. Entrepreneurs First functions something like an incubator, although Bentinck says EF thinks of itself more as a “talent investing studio.” It searches out individuals, usually with technical backgrounds, who also possess certain qualities related to pacing, productivity, determination, and even aggression, Bentinck says—qualities that alert EF that these individuals may outperform their peers. EF then guides them through the process of building a startup including helping them ideate if they don’t already have an idea and introducing them to potential co-founders. “We find exceptional individuals, pre-team, pre-idea, pre-company. Really all that we’re looking for is their entrepreneurial potential and then we run them through a process that helps them build a startup from scratch,” Bentinck says. The group that pitched this week included the top tier companies from EF’s European and U.S. programs. Each of these teams had been selected by EF and received $250,000 in pre-seed investment in exchange for 8 percent equity. “That’s the culmination of EF and we then send them off into the wild to build enormous companies,” Bentinck says. BY CHLOE AIELLO @CHLOBO_ILO

How AI Can Make You a Better Negotiator: A Step-by-Step Guide

Earlier this year, Jennifer Barnes received an email from a client in financial distress, asking to renegotiate their contract. As the founder and CEO of Optima Office, an outsourced HR and business services company based in San Diego, this was not the first time she’d received a message like this. She’s been in business since 2018, growing her 100‑employee company to around $18 million in revenue and earning a place on the Inc. 5000. There isn’t much she hasn’t seen. From experience, Barnes knew negotiating was going to burn the better part of an hour. First, she’d have to read the whole exchange. Then she’d think about how to respond. After that, she’d have to spend a lot of time writing and editing the response. She’d have to be diplomatic and keep her own emotions in check, she says: “Clients can be really unreasonable when they’re very low on funds.” This time, however, instead of working through it on her own, Barnes popped the email into her paid version of the AI chatbot Claude and asked it for a three‑point summary of the client’s demands. She then uploaded a brief synopsis of her perspective on the situation, did a light edit and hit send. Total time to craft the message that solved the problem? Five minutes. Negotiating is much of the work of growing a company. Whether it’s with clients, suppliers, investors, joint venture partners, contractors, or employees, as an entrepreneur it can feel like you’re constantly either preparing, actually doing, or managing the results of a negotiation. All of that is intellectually and emotionally demanding, says Emily DeJeu, professor at the Tepper Business School at Carnegie Mellon University, who teaches classes on negotiation. “Even in our textual exchanges, negotiation is happening as much with our guts as with our brains.” But with the advent of AI, using your own brain unassisted has become a bit passé. If sensitive, time‑consuming tasks like negotiating can be even partially off‑loaded without inadvertently blowing up your company, well, that’s pretty compelling. On the other hand, a recent MIT report found that 95 percent of companies are getting literally zero return on their generative AI investment. But it doesn’t have to be that way. By deploying today’s AI tools to prepare for negotiation—with a keen understanding of their current limits—you can leverage them to your advantage right now. In this Premium article you will learn: A step-by-step guide for incorporating AI into the negotiation process Best practices for combining human intuition with artificial intelligence How many hours a week you can expect to save by using AI When an hour takes five minutes For growing companies, there’s a wide array of AI tools from business-function-specific programs like Salesforce’s Einstein to generally available large language models like ChatGPT or Perplexity. All of these AI tools can analyze data and generate cogent text or other forms of output. Between the time savings and the added bonus of strictly controlling her tone, which might have had a sarcastic edge, Barnes says she quickly found herself relying on Claude for her frequent negotiation tasks: to research, prepare, and learn from prior wins. “It saves me about eight hours a week,” she says. “It’s like having another executive on the team.” Of course, that extra executive is sometimes fallible. “It makes mistakes. Don’t get me wrong,” she says. Generative AI systems are known to hallucinate, or produce inaccurate and even fabricated results. But AI seems just as confident when it’s wrong as when it’s right. That’s why Barnes says she would “absolutely not” allow an AI to send a message on her behalf without reviewing it first. When I asked Claude for a comment it agreed: “People getting real value are using me as a tool they control, not as a replacement for judgment. The moment anyone treats me as a peer negotiator rather than a research assistant, things get questionable.” Machine processing power versus human nuance Even if AI never made a mistake, there are serious questions about how well machines can accomplish the extremely human task of negotiating a complicated deal, says DeJeu, who is also hosting a conference on how businesses can use generative AI later this year. Artificial intelligence lacks the ability to read human emotion, which is often cited by the negotiators she’s trained as an advantage. DeJeu, however, disagrees. “Emotion has a distinct, powerful role to play in persuasion,” she argues. “Negotiation is one of the most human‑touch-necessary communication scenarios. It’s nothing but nuance.” Today’s tools, she says, aren’t capable of the key basic task of accurately reading a room. DeJeu acknowledges that not every negotiation is that deep. Not every supplier contract needs an analysis of subtle body language. And she certainly sees many benefits of using an AI tool to research and synthesize information ahead of a negotiation. “It can make you a little more nimble,” she says. DeJeu specifically finds using voice‑enabled AI for rehearsing negotiations to be beneficial. She suggests preparing 10 minutes for every 10 seconds of talking time in a negotiation. It’s hard to imagine a human critique partner enduring that without diminishing returns, but AI is tireless. Agents of the Future Of course, AI tools are rapidly evolving. Building on LLMs are the more eye‑catching AI‑powered agents—also known as agentic AI—which are capable of self‑direction. “AI agents act with autonomy and authority to find and negotiate deals with suppliers at scale,” explains Kaspar Korjus, CEO of Pactum AI. If technology continues its progress, AI agents will eventually handle every step of negotiation, from first contact to final contract and delivery. While going totally hands‑off is not an available option for most small to midsize companies today, corporate giants have been working on this for a while. For example, way back in 2021, Walmart worked with Pactum AI to create a pilot to handle certain supplier negotiations with AI chatbots. That led to a wider deployment in 2022—which you may recall is the year when ChatGPT first launched its public version. Now that ChatGPT has more than warmed up the general audience—the latest estimates for this one platform alone are 700 million users worldwide—agentic AI will only become more common. Nearly three‑quarters (72 percent) of chief procurement officers surveyed by Gartner say that AI tools like these are their top technology priority over the next five years. However, the next five years are not the next five minutes. “It’s still early days—we really don’t see any process that is fully agentic,” says Sesh Iyer, North America chair of BCG X, the tech build-and-design unit of Boston Consulting Group. For one thing, agentic AIs are still flummoxed by unexpected things, and that can make them take strange, if not brazen, shortcuts. For example, in a Carnegie Mellon study of a simulated company, researchers found that when an AI agent couldn’t find a particular person it needed to contact to complete a task, it just renamed another user. Overall, the top‑performing AI agent successfully completed only 24 percent of its tasks. But make no mistake—the tech is evolving fast, says Iyer: “We always overestimate what new technology will do in the short term and underestimate what it will do in the long term.” A practical playbook To avoid these problems of estimation in either direction, Iyer suggests starting with what you have right now. It’s easy enough to incorporate AI into negotiation prep with the tech your business probably already uses. Test it out with background research, brainstorming arguments and counterarguments, and use the voice function for rehearsal. But even at the most basic level of AI usage, it’s best to check in with your legal and IT security teams, since there are heavy privacy implications for both. When you’re using LLMs to their fullest for negotiation, you’re “allowing an incredible amount of access to information, including emails and alendars,” says Cameron Powell, co-founder of DeepLaw, a legal consultancy that uses AI tools for negotiation on behalf of its clients. Sharing this information can raise questions about confidentiality, liability, and intellectual property. When AI has proved its mettle and security to your satisfaction, the next step is using it to conduct deep analysis of your current contracts, sales, and negotiation processes. This could mean using AI to review your less-used suppliers that might be costing you more than they’re worth. You could also use AI to create side-by-side comparisons of your competitors or to provide a deep analysis of your current contracts to see what advantages you may be leaving on the table. AI can help manage tasks that are too time‑consuming or unwieldy to otherwise manage closely. Eventually, you can move to testing semi‑autonomous AI agents on repetitive or less nuanced negotiating tasks. Wherever you are in the process of integrating this new tool, Iyer suggests moving deliberately, and with a sharp eye toward integrating AI into your and your employees’ workflows. That MIT report that found little return on generative AI investment attributed much of the problem to enthusiastic focus on the gee-whiz technology itself, at the expense of truly considering how to make the best use of it today. “Focus on things that truly matter to your business,” urges Iyer. “Don’t try to do a thousand things at once.” BY ALISON J. STEIN

Meta’s Bold Strategy to Beat OpenAI Starts With These 8 AI Innovators

OpenAI might be the center of the AI development world these days, but the competition has been heating up for quite a while. And few competitors are bankrolled on the same level as Meta. With a market capitalization of more than $1.75 trillion and a CEO who’s not afraid to spend heavily, Meta has been on a hiring spree in the AI world for months, poaching top tier talent from a variety of competitors. It appeared recently that the wave of high-profile (and high-dollar) recruitments was coming to an end. In August, Meta quietly announced a freeze on hiring after adding roughly 50 AI researchers and engineers. This month, though, two more big names have joined the Meta roster. While Meta might have a gap to close with its AI rivals, the company has assembled an all-star team to catch up and move forward. Here are some of the most notable experts to come on board. Andrew Tulloch, co-founder of Thinking Machines Lab Tulloch partnered with OpenAI’s former chief technical officer Mira Murati to launch Thinking Machines Lab in February of this year. Now he’s returning to his roots. Considered a leading researcher in the AI field, Tulloch previously spent 11 years at Meta, leaving in 2023 to join OpenAI, then departing with Murati. Meta founder Mark Zuckerberg has been chasing Tulloch for a while, reportedly making an offer with a $1.5 billion compensation package at one point, which Tulloch rejected. (Meta has called the description of the offer “inaccurate and ridiculous.”) There’s no word on what Tulloch was offered that made him decide to move. Ke Yang, Senior Director of Machine Learning at Apple Yang, who was appointed to lead Apple’s AI-driven web search effort just weeks ago, is another big October Meta hire. At Apple, his team (Answers, Knowledge and Information, or AKI) was working to make Siri more Chat-GPT-like by pulling that information from the web, making his departure one of Meta’s most notable poachings. Meta convinced him to come over after recruiting several of his colleagues. Shengjia Zhao, co-creator of OpenAI’s ChatGPT Zhao joined Meta in June to serve as chief scientist of Meta Superintelligence Labs. Beyond co-creating ChatGPT, he also played a role in building GPT-4 and led synthetic data at OpenAI for a stint. “Shengjia has already pioneered several breakthroughs including a new scaling paradigm and distinguished himself as a leader in the field,” Zuckerberg wrote in a social media post in July. “I’m looking forward to working closely with him to advance his scientific vision.” Daniel Gross, co-founder of Safe Superintelligence As it did with Murati’s Thinking Machines Lab, Meta tried to acquire Safe Superintelligence, the AI startup co-founded by OpenAI’s former chief scientist, Ilya Sutskever. When that offer was rejected, Zuckerberg began looking for talent, luring co-founder and CEO Gross in June. Gross is working on AI products for Meta’s superintelligence group. By joining Meta, he’s reunited with former GitHub CEO Nat Friedman, with whom he once created the venture fund NFDG. Ruoming Pang, Apple’s head of AI models Pang was one of the first high-profile departures from Apple to Meta, making the jump in July. At the time, he was Apple’s top executive overseeing AI models and had been with the company since 2021. While there, he helped develop the large language model that powers Apple Intelligence and other AI features, such as email and webpage summaries. Matt Deitke, co-founder of Vercept Vercept is a start-up that’s attempting to build AI agents that use other software to autonomously perform tasks, something that caught Zuckerberg’s attention. Deitke proved hard to lure, though. He reportedly turned down a $125 million, four-year offer, but a direct appeal by Zuckerberg (and a reported doubling of that offer) convinced him to make the move (with the blessing of his peers). Kiana Ehsani, his co-founder and CEO, announced his departure on social media, joking, “We look forward to joining Matt on his private island next year.” Alexandr Wang, founder and CEO of Scale AI Wang left his startup to join Meta after the social media company made a $14.3 billion investment into Scale AI (without any voting power in the company). “As you’ve probably gathered from recent news, opportunities of this magnitude often come at a cost,” Wang wrote in a memo to staff. “In this instance, that cost is my departure.” Wang joined Meta’s superintelligence unit. Scale made its name by helping companies like OpenAI, Google and Microsoft prepare data used to train AI models. Meta was already one of its biggest customers. Nat Friedman, former CEO of GitHub Friedman was already a part of Meta’s Advisory Group before he was brought on full-time. That external advisory council provides guidance on technology and product development. Now, he’s working with Wang to run the superintelligence unit. Friedman previously was CEO of GitHub, a cloud-based platform that hosts code for software development. Most recently, he was a board member at the AI investment firm he started with Safe Superintelligence’s Gross. As for what Zuck is going to do with all this talent, the sky’s the limit, but there’s some catchup to do first. The Llama Large Language Model hasn’t quite matched up to those of OpenAI or Google, but with Meta’s gargantuan user base (3.4 billion people use one of the company’s apps each day), Meta’s AI could still be one of the most widely used in the years to come. BY CHRIS MORRIS @MORRISATLARGE

This Report Says AI Stole 17,000 Jobs This Year. The DOGE Effect Is Much Worse

AI evangelists continue to insist that AI is improving workers’ efficiency and thus business productivity, freeing up staff from mundane duties to do more meaningful work. Not as many boosters are cheering the fact that it’s just as easy for companies that have gone all in on the new technology to cut labor costs by replacing people’s jobs. According to a new report thousands of jobs have already gone from the job market this year as AI has assumed those duties instead, and fully 7,000 of the losses happened in September alone. All of this may feed into your thinking about rolling out AI at your own company. The data, from Chicago-based executive outplacement firm Challenger, Gray & Christmas, attributes 17,375 job losses to adoption of AI tech since the start of 2025. Most of these cuts were made public in the second half of the year, industry news site HRDive reports. The numbers are dramatic, especially since a similar report from Challenger in July said that among some 20,000 jobs lost to “automation” in the first half of the year, only 75 were directly connected to AI. Andy Challenger, senior vice president at the firm, told CFODive at the time that the suspicion was that many more jobs were actually lost to AI. “We do see companies using the term ‘technological update’ more often than we have over the past decade, so our suspicion is that some of the AI job cuts that are likely happening are falling into that category,” Challenger said then, also noting that some firms were being careful because they “don’t want press on it.” In the new report, Challenger noted that it’s mainly tech firms that are “undergoing incredible disruption,” because of AI. Challenger also backed up many earlier reports by noting that the buzzy, controversial tech is “not only costing jobs, but also making it difficult to land positions, particularly for entry-level engineers.” HRDive notes that it’s losses at Salesforce that may be linked to those massive AI-related job cuts in recent months, with Salesforce CEO Marc Benioff noting in August that customer service staff numbers were slashed by about 4,000 after AI agents took on some customer handling duties. The interesting wrinkle here is that Salesforce is one of the big tech names that is pivoting aggressively and openly to adopting AI tech, and is even selling it to its customers with the promise that agent-based AIs can save them money. Benioff in early 2025 also said “my message to CEOs right now is that we are the last generation to manage only humans.” In his vision for future company leadership, managers will be steering both AIs and humans through their day to day operations. While 17,000 jobs lost to AI sounds like a lot, it’s dwarfed by other causes, the Challenger report shows. DOGE-related actions is the “leading reason for job cut announcements in 2025,” the report notes, with 293,753 planned layoffs connected to DOGE activities, including reductions to federal workforce numbers and the cutting of contractor deals. Nearly 21,000 more jobs have been lost as part of what Challenger’s report says is “DOGE Downstream Impact,” where funding cuts have hit nonprofits that depend on federal grants. Traditional market and general economic concerns drove another 208,227 cuts in 2025, the report also notes. This means DOGE and the typical workings of the economy are responsible for around 30 times as many job losses than AI. But it would be unreasonable to assume AI’s body count won’t rise, considering Big Tech’s push to get AI into the workplace, while developing increasingly capable AI tools that can handle human jobs. And while Challenger notes that tech-centric firms are bearing the brunt of AI-related job cuts right now, it would be sensible to guess that other industries will soon follow. What’s the takeaway for your company? Primarily that it may be a good idea to reassure your staff that if you’re rolling out AI tools to streamline operations, you’re not actually planning on downsizing your workforce. ”AI won’t be stealing anyone’s job here” is a strong message that will build your team’s trust, assuming that this is actually the case. Another side effect may be a glut of workers in the job marketplace. Since many job seekers are using AI tools to boost their hunt for new employment, you may actually see many more applicants than before for open positions at your company, and your HR team may be quickly overburdened. BY KIT EATON @KITEATON

This Brooklyn-Based AI Company Just Raised $2 Billion to Compete With DeepSeek

A Brooklyn startup just raised $2 billion to build a rival to DeepSeek, the Chinese AI company. Called Reflection AI, the company is now valued at about $8 billion, up some 15-fold from last March, when it announced $130 million in funding. The company is less than two years old. Reflection, which launched in March 2024, originally aimed to build a “superintelligent autonomous coding system,” and use that as a jumping off point. Now, it is working on building an open alternative to the types of closed frontier models that giants like OpenAI are developing. In other words, Reflection wants to be the U.S. answer to China’s DeepSeek. “AI is becoming the technology layer that everything else runs on top of,” Reflection noted in a blog post about the funding. “But the frontier is currently concentrated in closed labs. If this continues, a handful of entities will control the capital, compute, and talent required to build AI, creating a runaway dynamic that locks everyone else out.” U.S. AI and crypto czar David Sacks praised Reflection on Thursday. “It’s great to see more American open-source AI models. A meaningful segment of the global market will prefer the cost, customizability, and control that open source offers. We want the U.S. to win this category too,” he posted on social media platform X. Aside from remaining globally competitive, Reflection says there are numerous benefits to frontier open intelligence, including safety, transparency, and accountability. (Frontier in this case refers to the most advanced, large-scale LLMs, like those currently in development behind closed doors at companies like OpenAI.) But it also flags the potential for misuse. High profile players in the space, like OpenAI’s Sam Altman, have publicly fretted about bad actors weaponizing AI; another concern is that others in the space are not putting in place adequate safeguards—even as Altman pushes to avoid regulation. OpenAI has since announced it is working on its own open model. “We believe the answer to AI safety is not ‘security through obscurity’ but rigorous science conducted in the open, where the global research community can contribute to solutions rather than a handful of companies making decisions behind closed doors,” Reflection’s blog says. The startup has spent the past year assembling a crack team of experts who have “pioneered breakthroughs including PaLM, Gemini, AlphaGo, AlphaCode, and AlphaProof, and contributed to ChatGPT and Character AI, among many others.” Its founders, Misha Laskin and Ioannis Antonoglou, worked on DeepMind’s Gemini and go-playing AI AlphaGo, respectively. The company also noted that it developed a large language model and “reinforcement learning platform capable of training massive mixture-of-experts (MoE) models at frontier scale.” TechCrunch reported that MOE models are a type of architecture that powers these super advanced, frontier LLMs. “We saw the effectiveness of our approach firsthand when we applied it to the critical domain of autonomous coding. With this milestone unlocked, we’re now bringing these methods to general agentic reasoning,” the blog states. Reflection also stated it has come up with a commercial model that will allow the company to sustain itself, while developing frontier models. It aims to release its first model early next year, TechCrunch reported. BY CHLOE AIELLO @CHLOBO_ILO

This Robotics Startup Just Emerged From Stealth With $300 Million to Create an ‘AI Scientist’

A new AI startup created by OpenAI and Google DeepMind alumni has emerged from stealth with $300 million in funding from some of the biggest names in tech. The company, called Periodic Labs, says it is fully dedicated to accelerating scientific discovery with AI. According to a New York Times story on Tuesday, the company was cofounded by Liam Fedus, one of the original creators of ChatGPT, and Ekin Dogus Cubuk, who led some of Google DeepMind’s materials and chemistry research teams. Cubuk’s team discovered 2.2 million new inorganic crystals, according to TechCrunch. The founders met while they were both working at Google, and connected over a shared desire to use large language models (LLMs) to advance the study of physics and chemistry. Unlike their former employers, both of which have in recent weeks released products that use AI to generate short-form videos, the founders say that Periodic Labs will focus entirely on using AI to test physics simulations. To do this, Periodic will open up a laboratory in Menlo Park that uses robots, powered by large language models, to run scientific experiments. According to Periodic’s website, their “goal is to create an AI scientist.” Robots will handle the same kind of physical experiments conducted by human scientists, but at a scale that’s humanly impossible. In an example given by the Times, an AI scientist “might run thousands of experiments in which it combines various powders and other materials in an effort to create a new kind of superconductor, which could be used to build all sorts of new electrical equipment.” Unlike humans, robots don’t need to eat, sleep, or take breaks from work, so they can run experiments for much longer. Eventually, according to the Times, the robots could learn which factors lead to success when trying to prove a hypothesis, and use that knowledge to improve their work. The founders think this process could result in breakthroughs for multiple industries, including semiconductors. That potential was enough for Fedus and Cubuk to raise a $300 million seed fund, led by a16z along with NVIDIA, Jeff Bezos, former Google CEO Eric Schmidt, and FIRST Robotics founder Jeff Dean. Over 20 top Silicon Valley researchers have left their jobs to join Periodic. In a video interview with a16z released alongside news of the fundraise, Fedus said that Periodic’s ideal customers are engineers and researchers in advanced industries like space, defence, and semiconductors. These engineers and researchers “don’t really have particularly good tools,” said Fedus, “and that is our opportunity. These are massive R&D budgets.” BY BEN SHERRY @BENLUCASSHERRY

Walmart’s CEO Just Gave a Sobering Prediction About AI. The Time to Prepare Is Now

Doug McMillon, as the CEO of Walmart, runs the largest private employer in the United States. When he talks about the future of work, it isn’t theory—it’s the lived reality of millions of families. In fact, more than 2.1 million people around the world get a paycheck from Walmart. That’s why it matters that, speaking at a workforce conference in Bentonville, Arkansas, last week, Walmart’s CEO didn’t mince words about artificial intelligence. “It’s very clear that AI is going to change literally every job,” McMillon said, according to The Wall Street Journal. “Maybe there’s a job in the world that AI won’t change, but I haven’t thought of it.” Look, a lot of people have predicted that AI will change the way we work in the future. For that matter, people are predicting that AI will change the way we do pretty much everything. It’s already changing the way we look for and process information. And it’s having a real impact on creative work, from generating ideas to editing photos. But this is different. This isn’t some kind of edge case where AI is doing something that benefits niche work. This is a sober assessment from someone who thinks about the livelihoods of millions of people, from truck drivers to warehouse workers and store managers. So far, much of the AI conversation around work has been about replacing humans with robots or computers capable of doing everything from menial tasks to coding. The pitch is that companies will save extraordinary costs as humans are replaced with AI that can do more work, faster, and cheaper. The fear among many employees is that automation will come for knowledge work the same way robots came for manufacturing. McMillon’s warning is different: AI isn’t confined to Silicon Valley jobs. It’s coming for the retail floor, the supply chain, the back office, and the call center. For example, AI can already predict what items a store will sell and when, automatically adjusting orders. That doesn’t eliminate the need for employees—but it will definitely change what their job looks like. McMillon also made another point: Walmart’s overall head count will likely stay flat, even as its revenue grows. That—if you think about it—isn’t just surprising, it’s incredibly revealing. The assumption is that AI equals fewer jobs. Instead, Walmart expects them to be different. To make that happen, the company is mapping which roles will shrink, which will grow, and which will stay stable. The strategy is to invest in reskilling so workers can move into the new jobs AI creates. “We’ve got to create the opportunity for everybody to make it to the other side,” McMillon said. This is the part of the warning many leaders ignore. Pretending AI won’t affect your workforce is irresponsible. Pretending AI only means job cuts is short-sighted. The challenge is to figure out what your workforce looks like and what you need to do to make the transition. There are a few reasons that Walmart’s perspective matters. The obvious one is because it’s the largest private employer in the world. It is the company that, single-handedly, affects the greatest number of people when it makes a change to its workforce. That’s why AI isn’t just a technology problem; it’s a leadership problem. It’s one thing for McMillon to say “AI will change every job.” It’s another thing to commit that Walmart will still employ millions of people, even if the jobs look different. He’s saying the responsibility to guide workers through change rests squarely on leaders’ shoulders. That’s a message worth hearing far beyond the company’s Bentonville headquarters. AI is often pitched as a productivity story. That’s true, but the bigger story is about people. Technology that changes “literally every job” also changes lives, families, and communities. The ripple effect is enormous when you’re a company the size of Walmart. By the way, Walmart isn’t perfect, but its approach offers a model. Instead of framing AI as cost-cutting, it’s framing AI as a transformation challenge. That may seem like semantics, but reframing the conversation makes all the difference between a fearful workforce and a resilient one. McMillon’s prediction is sobering precisely because it’s credible. He isn’t selling software or trying to impress investors. He’s planning for how millions of his own employees will navigate the AI future. If you’re leading a business—whether that’s 20 people or 20,000—the message is pretty clear. AI is going to change every job. Your job is to be thinking hard about what that means for your company. It means thinking about how it will impact your people and coming up with a plan. It seems like almost everyone agrees that AI will change almost everything about the way we all work. The only question is whether you’ll help your people prepare or leave them to figure it out on their own. By then, it will be too late. That’s why every leader should start now. EXPERT OPINION BY JASON ATEN, TECH COLUMNIST @JASONATEN

OpenAI wants to build the next era of the web, and it’s shelling out billions to do it

OpenAI was an artificial intelligence research lab little known outside of Silicon Valley before ChatGPT debuted in November 2022. Three years later, OpenAI has become synonymous with the AI boom, making it the envy of its tech peers and thrusting CEO Sam Altman into President Donald Trump’s orbit. ChatGPT writes apps, plans trips and browses the web on users’ behalf. And OpenAI is making inroads into shopping, entertainment, education and government services — laying out plans to become more like a platform than a basic app in its developer conference on Monday. As its software spills into more areas of online life, OpenAI is shelling out billions to become a leading player in the physical infrastructure for the AI future. With its latest major investment, announced on Monday, OpenAI will invest in 6 gigawatts of data center capacity powered by AMD chips. That deal follows similar agreements with Nvidia and Oracle. In some ways, OpenAI’s expansion is circular — it needs new applications to bring in the money to fund its massive computing power. And it needs even more computing resources to power those new tools. OpenAI’s rapid expansion comes against a challenging backdrop. Tech companies are competing fiercely to build the most powerful AI models, but some investors worry the market is in a bubble. What’s more, OpenAI is competing with tech giants such as Meta that already have vast tech ecosystems to help them expand and earn money from their AI tech. And OpenAI, which is not yet profitable, needs to find a way to continue raking in huge amounts of cash to fund its future endeavors. OpenAI did not respond to a request for comment on this story. ChatGPT: More than just a chatbot Google, Amazon and Meta laid the groundwork for the modern web by popularizing search engines, e-commerce and social media. OpenAI could do the same for the AI era by adding new capabilities to ChatGPT, which now has 800 million weekly active users, according to Altman. OpenAI wants users to get things done online without ever having to leave ChatGPT, which could one day put the app at the core of how people use technology, much like Apple’s iOS or Google’s Android system. Soon ChatGPT will be able to create user playlists directly on Spotify or browse apartment listings on Zillow right from chats, OpenAI announced on Monday. In late September, OpenAI launched a tool called Instant Checkout that lets users buy certain items directly through ChatGPT. ChatGPT also now has a study mode, which tailors prompts and responses for students using the tool for schoolwork. And its new Sora 2 app is challenging Meta and TikTok with a scrollable feed of AI-generated short-form videos. OpenAI could even challenge the most prominent device in consumers’ daily lives: the smartphone. The company is collaborating with former Apple design chief Jony Ive on a new AI hardware product, though details are slim. (OpenAI’s peers like Google and Meta are chasing hardware markets by releasing smart glasses with built-in AI assistants.) OpenAI’s trajectory mirrors the rise of Google parent Alphabet, which built its business around indexing the web and now has a foothold in everything from consumer tech devices to health research. Thomas Thiele, an AI expert at management consulting group Arthur D. Little, said he sees similarities between the two companies. Google “has become this very broad corporation that has an inevitable footprint in everything we see on the internet,” Thiele said. “OpenAI is also aiming for a much bigger footprint.” Billions on data centers But scaling up those AI efforts means investing heavily in the sprawling data centers and infrastructure necessary to power them. OpenAI is shelling out billions of dollars to build a massive physical footprint, with plans for AI data centers across the United States and around the world. “We need as much computing power as we can possibly get,” OpenAI President Greg Brockman told CNBC on Monday. In January, the company announced a partnership with Oracle and SoftBank to invest up to $500 billion in a company called Stargate to build more AI infrastructure in the United States. The group’s first project, a one-million-square-foot data center, is already under construction in Abilene, Texas, with additional sites planned in Texas, New Mexico and the Midwest. OpenAI agreed in July to pay Oracle another $300 billion over five years to develop additional data center capacity for Stargate. Last month, OpenAI said it would buy enough Nvidia AI chips to power 10 gigawatts of data center capacity in exchange for a $100 billion investment from the chipmaker. And while analysts expect Nvidia — the undisputed leader in AI chips — to remain OpenAI’s core infrastructure partner, the ChatGPT maker is now also hedging its bets with its AMD deal. OpenAI has also signed onto partnerships to build out AI infrastructure abroad, including in the United Kingdom and United Arab Emirates. OpenAI’s aggressive expansion could be critical to keep up with rivals like Meta, Microsoft and Google that have spent decades building their digital ecosystems, said Daniel Keum, an associate professor at Columbia Business School. Google, for example, has the advantage of plugging its AI into popular services like Gmail and Google Docs. “ChatGPT is great right now, but it’s not ChatGPT versus Copilot. It’s ChatGPT versus the Microsoft bundle,” said Keum. So for OpenAI, working with chipmakers to maintain the most advanced large language models could give it a leg up, he said. But to carry out its ambitious infrastructure plans, OpenAI needs to continue bringing in a whole lot of cash. The company is reportedly valued at $500 billion. But it’s still far from profitable; it posted an operating loss of $7.8 billion in the first half of 2025 and is still ramping up data center spending, according to a report from tech news site The Information. It’s unclear whether OpenAI’s bid to turn ChatGPT into an all-encompassing platform will put it on the path to profitability. William Lee, a corporate investor at SuRo Capital, sees it as a “chicken-or-the-egg” problem, he said in an interview with CNN. Demand may be hard to gauge ahead of time, but the more OpenAI customizes ChatGPT for tasks like shopping and schoolwork, the more people could use it for those activities. It’s a strategy that has worked for the tech giants of today — spend aggressively to make your technology essential to millions of users’ lives, figure out how to make money from them later. OpenAI is clearly betting that it will pay off again. “AI revenue is growing faster than, I think, almost any product in history,” Brockman told Bloomberg. “At the end of the day, the reason this compute power is so important, is so worthwhile for everyone to build, is because the revenue ultimately will be there.”

Need a Social Media Influencer for Your Brand? There’s an AI for That

Influencer marketing campaigns can be powerful tools, helping brands acquire new customers, make product launches go viral, and even drive growth in down markets. But identifying which content creators you should partner with is often a time-consuming and complicated task. With the launch of its new AI-powered creator discovery tool, influencer marketing platform Superfiliate aims to change that. The Venice, California-based company, which was founded in 2021 by Anders Bill and Andy Cloyd, uses first-party data from Meta to match businesses with influencers, according to a press release. “Instead of manually scrolling through platforms hoping to find creators who might work, brands can now leverage the same recommendation intelligence that makes Netflix, Spotify, and even your Instagram Explore page so effective,” CEO Cloyd said in the press release. Marketers can use the tool to search for specific kinds of content creators, conduct research on their content style and past brand partnerships, determine if they’re brand safe, and contact them directly through email. A promotional video by Superfiliate, for example, shows that a brand-side user can type in something like “Find me home decor creators with 50K+ followers” in the tool’s search bar and surface several viable options. They can also upload a creator’s social media handle and find several influencers with similar niches and followings, per the video. There are, of course, plenty of tools already on the market that help brands search for creators, such as those by CreatorIQ and Grin. What makes Superfiliate’s stand out, according to Bill, who now serves as the company’s chief product officer, is its direct partnership with Meta. The social media giant’s participation “enables platform-native infrastructure that’s fundamentally more accurate and compliant than scraping-based approaches,” he said in the press release. While Superfiliate doesn’t publicly share its prices, Cloyd previously told Inc. that his company charges brands for the use of its platform plus an additional fee based on either their total ad spend or the upside of sales they generate through Superfiliate. If a business wants to use only specific parts of the platform, it’ll charge a flat rate instead. BY ANNABEL BURBA @ANNIEBURBA

This Company Says 1 New AI Feature Can Handle 20 Hours of Work in Seconds

Popular wedding planning platform The Knot has released an update to its mobile app that uses AI to streamline the process of finding local vendors. In a press release, the company said that the new update “cuts over 20 hours of planning work to just seconds.” The reimagined “planning experience,” as The Knot calls it, allows couples to browse through thousands of photos of weddings to create a vision board. By clicking an icon, users can activate a new feature called “make it yours,” which scans the image and then searches through The Knot’s database of venues and vendors to find similar options that “fit your vibe, budget, and location.” Christine Brown, The Knot’s VP of product, says that the company built this new AI feature entirely in-house, rather than relying on AI models from external providers like OpenAI or Anthropic. To create the feature, Brown says the company trained its own models on “more than a million images accessible on The Knot.” To test its effectiveness, The Knot ran a two-month pilot in which thousands of couples were given early access to the tool. As an example of how the new feature can help amateur wedding planners save time, Brown pointed to one of the most time-consuming aspects of throwing a wedding: picking a venue. Brown’s team estimated that most couples take roughly six weeks to pick a venue, spending 3.5 hours per week devoted to the search. That adds up to 21 hours of total searching time, which Brown says can now be reduced to minutes thanks to this new tool. The Knot says that this update is just the first step in a larger push to introduce AI-powered wedding planning features. As for what’s next, she says the company is building AI tools to help both couples and professional wedding planners and vendors. One of those tools is an AI-assisted email reply feature that allows vendors to convert more leads into bookings. “We see AI as a powerful force to support the planning journey,” Brown says, “helping couples and vendors save time, while still keeping personalization and human touch at the heart of the wedding experience.” BY BEN SHERRY @BENLUCASSHERRY

Microsoft Is Adding Anthropic’s Claude to Its AI Tools. Here’s What It Can Do for Businesses

Microsoft is expanding the lineup of AI models used to power 365 Copilot, its workplace-focused AI service. The move is a sign that Microsoft is actively working to lessen its reliance on OpenAI‘s models after investing over $10 billion in the company. In its blog post announcing the news, Microsoft said that while 365 Copilot will continue to be primarily powered by OpenAI’s models, users will now be able to harness Anthropic’s models in two specific ways. One is in Researcher, a 365 Copilot feature that searches the internet and analyzes internal data like emails, Teams chats, and files, in order to conduct deep research. Normally, Researcher runs on models developed by OpenAI, but 365 Copilot customers will now have the option of using Claude’s Opus 4.1 model (the company’s most advanced model currently available) instead. Microsoft said that Opus 4.1 in Researcher could be used to accomplish tasks like “building a detailed go-to-market strategy, analyzing emerging product trends, or creating a comprehensive quarterly report.” The other method for using Claude in 365 Copilot is within Copilot Studio, a feature that enables users to build customized AI agents that can automate workflows. Users will now be able to easily select Claude Opus 4.1 or Claude Sonnet 4 (Anthropic’s mid-sized model) when creating agents. Microsoft says users will even be able to orchestrate whole teams of agents, all powered by different AI models, to work in tandem in order to accomplish tasks. Workplaces with Microsoft 365 Copilot licenses can now use Claude in Researcher and Copilot Studio, but only if opted-in by an administrator. Microsoft wrote that “this is just the beginning,” and that users should stay tuned for Anthropic models to “bring even more powerful experiences to Microsoft 365 Copilot.” Microsoft is also reportedly working on an AI marketplace for news and media publishers, according to Axios. The marketplace would enable publishers to sell their content to AI companies, who would in turn use that content to train their new AI models. Axios reported that Microsoft discussed plans for the marketplace at its invite-only Partner Summit in Monaco. BY BEN SHERRY @BENLUCASSHERRY