• Silvio Pfeufer uses Saltz to save time on food procurement at his Michelin-starred restaurant, Matthias.
• Saltz uses agentic AI to modernize a typically complicated process between food buyers and sellers.
• For independent eateries, the streamlined process can reduce costs and increase access to high-quality ingredients.

When Silvio Pfeufer first got into cheffing, he was surprised by the amount of administrative work involved.

"It's not only the evening, sending nice plates to guests. There's a lot of stuff to do to make that happen," Pfeufer told Business Insider. He took long phone calls from food suppliers and producers that kept him away from the kitchen.

Sourcing and buying food is a process riddled with disorganization, since farmers, wholesalers, logistics providers, and restaurants communicate in different ways — by phone, email, text message, and PDF blast — starting as early as 4 a.m. The cost and availability of products are constantly in flux, so sellers issue frequent updates: The price of wild-caught fish could change three times a day, for example.

When Pfeufer opened his own restaurant in 2024, he wanted to streamline his food procurement process. At Matthias, which pays homage to his late grandfather, Pfeufer uses an AI-assisted platform for food business owners called Saltz to speed up food procurement and reduce phone calls.

The time savings are critical, said Pfeufer, a co-owner and head chef at the Berlin-based eatery. Using AI agents, Saltz connects restaurants directly to suppliers via a marketplace that brings together disparate catalogs, transactions, and logistics so restaurants can compare and buy fresh, high-quality, and specialty products more quickly than with traditional food procurement processes.

The platform's standardized, real-time food data can be a game-changer for independently owned and operated restaurants like Pfeufer's, given their limited purchasing power. Saltz said that thousands of buyers and hundreds of suppliers use its technology, though it didn't share exact numbers. It said around 80% to 90% of its buyers are independent restaurants.

Connecting with high-quality and specialty suppliers

Pfeufer said he chose Saltz because he liked how the platform modernized old processes. In the two years he's used it, it's allowed him to discover new suppliers, order outside business hours, and gain better oversight on pricing, he said.

Founded in 2022, the startup uses AI agents to ingest PDFs, emails, and text messages that food sellers send to Saltz, standardize and enrich all the data, and consolidate it into a single platform. Its AI agents automatically update each seller's listings on the platform, eliminating the need for manual updates.

Buyers, meanwhile, see once-disparate product options, information, and up-to-date prices in one place. They can also order food at a time that's convenient for them and track their deliveries. "On a Sunday, at night, or in the morning, I can do it by myself, and don't have to have all these calls," Pfeufer said.

At Matthias, which was awarded its first Michelin star in 2025 after 10 months of service, meals must meet a high standard every day, and products have to be of the freshest quality. Pfeufer orders vegetables, milk, and other items from Saltz every week, plus fresh fish twice a week. Before using Saltz, the quality of his products wasn't necessarily worse, he said, but access to new or specialist suppliers was more limited and supply chains were longer.

"The fish is now often sourced directly from the trader or farmer, without the need for intermediate storage. This allows us to avoid additional storage times that could negatively affect freshness," Pfeufer said, adding that more oversight into the supply chain — and it being shorter — means his food is more consistently high quality.

"We've definitely been able to connect with better suppliers," he told Business Insider. The chef added that he still works directly with certain local farmers, as he is often on the hunt for rare items that aren't on the platform.

For Pfeufer, using Saltz also allows him to see more costs upfront, so he can better plan how much to charge for new dishes. "It makes all the calculations much easier, which is very important for us," he said.

Agentic AI acts as a foundational tool

Saltz was founded by brothers Andrius and Thomas Šlimas, who previously built the Shopify-acquired dropshipping platform Oberlo. After spending four years inside Shopify's supply chain machine, the brothers teamed up with industry veteran Reinis Štrodahs. Their goal: modernize the $9.8 trillion food procurement industry.

"It's impossible to make sense or structure that chaos of information which lives in different places and has no common structure," Andrius said.

The Šlimas brothers said that previous unsuccessful attempts to modernize food procurement took two approaches: either trying to force suppliers and buyers onto a single platform, which required them to change how they work, or taking on the time-consuming task of manually inputting every PDF, email, and text message.

With Saltz, AI agents upload and update product listings for sellers, so individual stakeholder workflows don't have to change. Making agentic AI foundational to the process, Tomas said, gives their platform an edge.

"That gives us a speed advantage, and in this market, speed compounds into market share."

• Around 600 Google employees urged CEO Sundar Pichai to reject classified Pentagon AI deals.
• They said they want to see AI benefit humanity, not be used for autonomous weapons or surveillance.
• Google and the Pentagon are in talks to use Gemini in classified settings, per a recent report.

Around 600 Google employees sent a letter to CEO Sundar Pichai on Monday, urging him not to let the company's AI technology be used by the US military for classified operations.

The letter, signed by employees in Google's DeepMind and Cloud divisions, cited a recent Information report that Google and the Pentagon were negotiating the use of Google's Gemini AI in classified settings.

"As people working on AI, we know that these systems can centralize power and that they do make mistakes," the employees wrote in the letter. "We feel that our proximity to this technology creates a responsibility to highlight and prevent its most unethical and dangerous uses."

"Currently, the only way to guarantee that Google does not become associated with such harms is to reject any classified workloads," employees continued in the letter. "Otherwise, such uses may occur without our knowledge or the power to stop them."

Google didn't immediately respond to a request for comment from Business Insider. Google has not yet responded to the letter, said Jane Chung, the founder of Justice Speaks, a communications firm representing the workers. Bloomberg first reported on the letter.

Google has long faced internal pushback to its efforts to work with the US military. In 2018, it decided not to renew Project Maven, a Department of Defense contract to integrate AI into military operations, following pressure from hundreds of employees. Palantir later picked up the deal.

The same year, Google established a set of AI principles, including a pledge not to use AI for weapons or surveillance. Last year, it updated those AI principles to remove wording around weapons and surveillance.

The company also secured new contracts with the Pentagon last year to use its AI and cloud products. In March, the company said it would provide the Pentagon with AI agents in a non-classified setting. It also told Google DeepMind employees during a January meeting that they should expect more of these types of deals.

In the letter, Google employees raised concerns that classified work would lead to a lack of oversight into how the company's technology is used.

"We want to see AI benefit humanity; not to see it being used in inhumane or extremely harmful ways," the employees wrote. "This includes lethal autonomous weapons and mass surveillance but extends beyond."

Read the full letter below:

Dear Sundar,

We are Google employees who are deeply concerned about ongoing negotiations between Google and the US Department of Defense. As people working on AI, we know that these systems can centralize power and that they do make mistakes. We feel that our proximity to this technology creates a responsibility to highlight and prevent its most unethical and dangerous uses.

Therefore, we ask you to refuse to make our AI systems available for classified workloads.

We want to see AI benefit humanity; not to see it being used in inhumane or extremely harmful ways. This includes lethal autonomous weapons and mass surveillance but extends beyond. Currently, the only way to guarantee that Google does not become associated with such harms is to reject any classified workloads. Otherwise, such uses may occur without our knowledge or the power to stop them.

Making the wrong call right now would cause irreparable damage to Google's reputation, business, and role in the world. At this very moment, the safety of our own workforce and critical infrastructure are under active threat. Human lives are already being lost and civil liberties put at risk at home and abroad from misuses of the technology we're playing a key role in building.

We know from our own history that our leaders can make the right choices, for ourselves and for the world, when the stakes are high.

Today, we call on you, Sundar, to act according to the values on which this company was built, and refuse classified workloads.

• I asked ChatGPT to identify the unnamed male celebrity who allegedly tried to sext Lena Dunham in 2012.
• It told me it was Don Rickles, which I feel pretty certain is not correct.
• So what are we doing here, folks? Learning to use AI?

Did you hear about the time Don Rickles tried to chat up Lena Dunham in the middle of the night?

No? Let me explain. First, we need to talk about Reese Witherspoon.

See, I'm a simple woman. I have only two interests: tech news and celebrity gossip. So I was naturally intrigued by a recent online fuss over Reese Witherspoon's admonition for women to learn to use AI. It sparked so much backlash that she had to issue a follow-up explanation.

I've also been intrigued by Lena Dunham's new book. (They're related — sort of. Keep reading!)

I think Reese is generally right about AI — she's saying the same thing that every other business leader is saying. But her comments did make me think a little more about what "Learn to use AI" even means. Writing emails with ChatGPT? Understanding the technology behind different models? Vibe coding? What level of "using AI" is expected here to stave off falling behind in the workforce and life in general?

One area I've really leaned into is using ChatGPT as a sort of super Google — to find something I know is online but would take some effort to dig up with a normal search engine.

A recent example? It's related to — of course — celebrity gossip.

I was reading Dunham's new memoir, "Famesick," which is full of moderately juicy celebrity gossip about named people and also blind items — celebrity gossip that gives a few clues about the identity of the person without naming them, a fun little riddle for the readers to solve.

One blind item is about an unnamed male celebrity who — allegedly — sent Lena a flirty late-night text message after meeting her backstage while taping "The View" in 2012. I figured I could solve this blind item by finding out who the other guest was on the same episode — information that should be online somewhere, but would take me forever to find.

So I asked ChatGPT to identify the male guest on "The View" episode that Lena was also on that year. At first, ChatGPT told me that it was only the four female cast members from the show. When I asked again who the other male guest was, the suggestions were Chris Evans and Chris Hemsworth. (Not so. They appeared on a separate episode that same year, according to IMDb.)

That time Don Rickles chatted up Lena Dunham

When I said, "No, a comedian," as Dunham had described the man, ChatGPT confidently provided a new answer: It was legendary comedian Don Rickles who'd texted Dunham after the show.

I laughed out loud because of all the possibilities of who sent a late-night "u up?" text, I feel fairly certain it was not Don Rickles, who would've been 85 years old at the time.

Dunham's description of the man: "a bit of an American Hugh Grant, famous for that sort of chattery charm and his ability to woo his onscreen paramours with his fast-talking, hand-flapping anxiety. Ostensibly a comedian, he was there to promote his Gothic-tinted movie, where he had made a dramatic turn." Doesn't exactly sound like a Borscht Belt insult comic Don Rickles to me.

After spending way too much time searching the internet for answers on this — the old-fashioned way — I can make some guesses about how and why ChatGPT was so wrong here. IMDb's episode guide for episodes of "The View" from 2012 is spotty, with entries for some episodes missing information about guests, and no accessible video clips online. The only proof I found that Lena Dunham ever appeared on "The View" on April 20, 2012, was a Vulture blog post from that day, complete with an embedded YouTube clip that has been marked private.

Knowing this, I can start to see how AI got confused: When there's a lack of information, AI sometimes blurs together what it can find to try to spit out a plausible answer. Chris Evans and Chris Hemsworth appeared on the May 4, 2012, episode of "The View," and Dunham and Rickles appeared together on an episode in 2016.

ChatGPT doing this kind of thing — basically, taking a guess at what you might want to hear — could be useful if you're trying to write an email to a friend, maybe? It's not useful, obviously, if you're looking for a specific fact and it just plain makes something up.

For the record: Neither Lena nor Don (who died in 2017) nor the National Comedy Center, which is the keeper of the Rickles archive, responded to my requests for comment.

Are we stuck in a pizza glue loop?

Look, I get it. It's not particularly exciting to point out that ChatGPT gets things wrong in the spring of 2026. We know this, or at least we all should know this. Still, I keep coming across so many obvious mistakes when asking AI for factual things. These are the glaring mistakes I catch when I know that what AI has generated is not the right answer.

But what about the mistakes that I don't catch — or don't even know to catch? Things that I blindly accept as fact? For work-related stuff, I'll always double-check, but in those cases, am I actually saving myself any time?

How soon will this improve? Will we be stuck in a pizza glue loop forever? Is this what's going to make a bunch of lawyers and tax CPAs lose their jobs? I mean, OK, sure.

Here's where Witherspoon's and other bosses' idea of "Learn to use AI!" feels frustrating. I feel fairly confident about using various AI tools and have a decent concept of how they work. I am a woman, and I have learned to use AI! And yet, here I am, still unsatisfied.

There's a gap between what Reese Witherspoon wants for me and what I want out of AI — and the wholesome image of comedy legend Don Rickles. For now, those things just aren't lining up right.

• Meta has set goals for some employees on how much they should use AI.
• They include targets for using AI code assistants, agents, and other tools.
• Meta CEO Mark Zuckerberg has said he wants the company to be "AI-native."

Mark Zuckerberg wants Meta to be "AI-native." An internal document shows one way the company's CEO plans to get there.

The company has set goals for how much some employees should use AI tools for tasks such as coding.

Meta employees created a document to collect information about these goals from across different organizations, according to a copy seen by Business Insider. It includes goals set late last year and for 2026.

Tech companies are using various methods to motivate staff to use AI, such as tying AI use to performance reviews and gamifying AI use with competitive leaderboards.

The document states that Meta's creation org, which is responsible for building and maintaining core creative experiences, set a goal for the first half of 2026 that 65% of engineers are expected to write more than 75% of their committed code using AI. Committed code is code that has been saved and tracked in a project.

Meta's Scalable Machine Learning org, which focuses on AI models and infrastructure, had a goal for February 2026 to achieve 50% to 80% AI-assisted code, the document said. It cited a comment alongside this goal from a senior engineering manager that said: "We are not tracking this via metrics."

The document also listed several companywide goals for Q4 2025 for central products — a horizontal org spanning Messenger, WhatsApp, Facebook, and other major products. One target is for 80% of mid to senior-level engineers to adopt AI tools such as DevMate, Metamate, and Google's Gemini, with a note that the focus is on "tool adoption" rather than the percentage of code written by AI.

It said that 55% of code changes from software engineers across the central product orgs should be "Agent-Assisted."

It is not clear whether the goals listed in the document are tied to performance reviews.

"It's well-known that this is a priority and we're focused on using AI to help employees with their day-to-day work," a Meta spokesperson told Business Insider. They said that Meta's performance program is focused on rewarding impact from AI tools, not just usage.

Here's a breakdown of Meta's goals in the memo:

• Companywide Q4 2025 Goals (Central Products)

  • 55% of software engineers' code changes should be agent-assisted.
  • 80% of mid to senior-level engineers should adopt general AI tools.

• Scalable ML Team Goal (Feb 2026)

  • Target: 50% to 80% AI-assisted code.

• Creation Org H1 2026

  • 65% of engineers should write more than 75% of their committed code using AI.

(Note: Some technical terms have been rephrased for clarity)

Mark Zuckerberg's AI odyssey

Zuckerberg is aggressively trying to make Meta what he has called an "AI-native" company. Meta has started tying employee performance to their AI usage, Business Insider reported last year, and staff are using Meta's internal AI bot to write reviews for their peers.

More recently, the company rebranded some employees within a division of Reality Labs with one of three titles: "AI Builder," "AI Pod Lead," or "AI Org Lead."

The change comes as Meta is adopting smaller teams and moving toward a flatter organizational structure.

"Our ultimate goal is to drive a step change in engineering productivity and product quality," read a memo about the changes, which was reviewed by Business Insider. "To achieve this, we're fundamentally rewiring how we operate, how we are structured, and how we support each other."

Andrew Bosworth, Meta's CTO, told staff on Tuesday that he would take charge of Meta's "AI for Work" initiative, which is designed to boost the company's internal adoption of AI tools, according to a memo reviewed by Business Insider and first reported by The Wall Street Journal.

Meta laid off several hundred employees across Reality Labs and other orgs this week.

Have something to share? Contact this reporter via email at hlangley@businessinsider.com or Signal at 628-228-1836. Use a personal email address and a non-work device; here's our guide to sharing information securely.

• JPMorgan software developers say the bank is raising its expectations for AI use.
• Internal company communications reveal the bank's new AI targets.
• The updated objectives affect members of its global developer workforce.

JPMorgan Chase's message to its global armada of software developers is clear: embrace AI or risk falling behind.

Internal company documents seen by Business Insider and posted to JPMorgan's intranet for employees lay out a series of new expectations for the bank's software engineering workforce, who comprise the majority of its 65,000-person-strong Global Technology division. The newly listed objectives, published on the intranet earlier this month, say all software and security engineers are expected to "drive excellence" by adopting AI and "contributing to initiatives that improve productivity, speed, scalability, and impact."

One document authored by the bank's human resources leaders laid out two core objectives for software engineers: step up their coding game, and start harnessing AI to save time and get more done. The new language about objectives "will be added automatically and will appear by the end of March," an image of the document on the intranet showed — a reference to upcoming changes to employees' goals expected to take effect at the end of this month. The firm also instructed workers to develop clear goals with their managers that align with the bank's new objectives.

"Demonstrate measurable improvement in code quality, speed and productivity through regular use of approved AI coding assist tools, contributing to the team's overall efficiency targets," read one goal written by HR. "Engage in identifying, implementing and optimizing AI-driven automation opportunities within technology lifecycle management (TLM) processes to drive efficiency and support capacity unlock initiatives, ensuring all enhancements leverage current technology assets before considering new solutions."

A spokeswoman for JPMorgan declined to comment.

JPMorgan is among Wall Street's biggest spenders on technology and artificial intelligence, with projected tech investments reaching roughly $20 billion in 2026 — far exceeding peers like Goldman Sachs. Across corporate America, companies including Meta and Google have begun pushing employees to adopt AI tools and, in some cases, evaluating their use.

Business Insider spoke to five engineers across the bank who said the push to adopt AI has been felt far and wide — in managerial conversations, in intranet posts, and through dashboards that display who's using certain AI tools, and who's not. They added that discussions about productivity and AI adoption have become more frequent in recent weeks. It all comes as developers get ready for a pilot of Anthropic's Claude Code to be rolled out as soon as April, said a longtime IT developer in the Global Technology group. Claude Code would be made available alongside the four other large language models coders are already using: two from OpenAI's ChatGPT, and two from Anthropic's Claude.

'Anxiety' among developers

The developers Business Insider spoke to said they've been encouraged to use AI tools for a wide range of tasks, from writing code to preparing presentations. One dashboard that tracked adoption and usage of the bank's GitHub Copilot appeared to show details as granular as which employees had installed it and identified individuals as "light," "heavy," or "non" users.

For some, the message has added pressure inside a firm that has drawn scrutiny in recent years for its use of internal monitoring tools and performance tracking. Business Insider published a series of reports on the firm's Workforce Activity Data Utility in 2022, a program that collected data points about how employees were spending their day — from the length of video calls to how long they spent drafting emails to where they were sitting in the office.

"There's a lot of anxiety in the environment right now," the longtime IT developer said. Those who don't use AI risk being seen as underperforming, the developer said. Another developer said their manager said in a recent meeting that availability of the new AI tools comes with an "expectation" that velocity and output should show "a noticeable increase" quarter over quarter.

Three of the five developers Business Insider spoke to said the tools are helpful, despite discomfort over the tracking.

New performance dimensions

The updated guidance on AI use comes as the bank implements other adjustments to how it ranks workers' success on the job. Going forward, the bank said on the intranet portal, it's streamlining some of the primary "dimensions" it uses to grade employees, pivoting to using two categories: "what you achieve" — business outcomes — and "how you achieve it," including adherence to the firm's behavioral principles.

According to screenshots from the bank's intranet, JPMorgan will segment workers into three buckets: "stand out" for those who exceed job standards, "achiever" for the majority of employees, and "needs improvement" for those who require "additional support" and have struggled to perform consistently.

Another page Business Insider reviewed listed skills non-managers working in software engineering were expected to display across "all performance dimensions." One is "Data Fluency," noting that the skill is applied by those who develop and drive "adoption of new tools or methodologies to leverage data in the flow of work." "Rate of adoption" is cited as one measurement of the employee's impact toward exhibiting the skill in practice.

The documents from the JPMorgan intranet echo the firm's long-standing culture of internal monitoring and data collection, making clear that continuous performance tracking is vital for keeping workers on target throughout the year.

"You and your manager will use your objectives to track your progress during the year, recognize impact, and streamline your annual review," the firm wrote on an internal page tied to goals.

Have a tip? Contact these reporters via email at ralexander@businessinsider.com or SMS/Signal at 561-247-5758 or atecotzky@insider.com or Signal at alicetecotzky.05. Use a personal email address, a nonwork WiFi network, and a nonwork device; here's our guide to sharing information securely.

• In October, Sam Altman said "erotica for adults" was coming to ChatGPT.
• Now those plans are reportedly being mothballed.
• It's fine for young startups and even mature companies to try out new ideas. But OpenAI and Altman are trying out a lot.

Last fall, Sam Altman told us he was about to bring spicy chat — "erotica for adults," in his words — to ChatGPT.

That never happened, and now it looks like it never will: Altman's OpenAI has put those plans on hold "indefinitely," per the Financial Times.

This is Altman's second big walkback in the last few days. Earlier this week, the company canned Sora, the briefly popular video app it rolled out last fall. I've asked the company for comment.

Both retreats are supposed to be part of a new push at OpenAI to focus the company's efforts on things that could make money today, as it preps for an IPO at the same time it faces real competition from the likes of Google and Anthropic.

So all this starting and stopping could be viewed as necessary growing pains at a fast-growing tech company — ones that won't mean anything in the long run, if it delivers on its world-changing ambitions.

Not only that, but Altman told us we should expect this sort of stuff. "Please expect a very high rate of change from us," he wrote last fall, after hearing from content owners who were outraged to find their stuff on Sora without their permission. "We will make some good decisions and some missteps, but we will take feedback and try to fix the missteps very quickly."

It's not that companies aren't allowed to make wrong turns and head up dead ends as they grow up, and even once they're fully mature. That kind of pivoting is celebrated in tech (and is why very few people are mad that Mark Zuckerberg has stopped telling us the metaverse is the future, or that Google once bought Motorola and decided that was a bad idea a couple years later.)

But "move fast and break things" lands differently when the company doing the moving and breaking isn't running a photo app or playing around with crypto.

Instead, OpenAI and its competitors say they're leading us into a world where everything — the way we live and work (or don't work) and fight wars and everything else — will change in fundamental ways.

And investors have bought this pitch, which means our economy now seems yoked to all this — which means all of us are yoked to it, even if we never touch a chatbot.

Which makes me slightly queasy to see Sam Altman promise dirty chats in October, and then walk away from the plan less than six months later.

Not because dirty chat is obviously absurd. Lots of people in AI think romantic or sexual chatbot conversations are a real use case and could be a real business.

But the reasons it might be a bad idea for OpenAI were pretty obvious from the start. It's a giant, heavily scrutinized company that wants to be treated as central and indispensable, and it's only going to get more scrutiny.

If those objections only became real after Altman floated the idea in public, that's not charming startup experimentation. It's a sign that OpenAI is still making itself up as it goes. And that would be easier to shrug off if the rest of us weren't already being told to build our lives, jobs, and businesses around what OpenAI says comes next.

• A large division within Meta Reality Labs is undergoing an overhaul to become fully "AI-native."
• The unit is now organized into "pods" made up of "AI builders" and "AI pod leads."
• This new push and the latest layoffs at Reality Labs are unrelated, Meta said.

Meta is rebranding some employees as "AI builders" and organizing them into AI-native "pods," according to a leaked memo obtained by Business Insider.

The memo described an overhaul of roles, titles, and team structures across a 1,000-employee team within Meta's Reality Labs. It's part of a broader, aggressive push by Meta to adopt small teams and use AI.

The pilot program was announced last month within the Reality Labs team that builds developer tools. Everyone in the division will now have one of three titles: AI Builder, AI Pod Lead, or AI Org Lead. That's to encourage a shift toward a flatter organization, a structure that Meta CEO Mark Zuckerberg has advocated.

"Our ultimate goal is to drive a step change in engineering productivity and product quality," the memo reads. "To achieve this, we're fundamentally rewiring how we operate, how we are structured, and how we support each other."

When asked for comment, Meta referred Business Insider to comments earlier this year from Zuckerberg that 2026 is the year AI will begin to "dramatically change the way we work," with projects that once required large teams potentially handled by one, "very talented" person.

According to the memo, each pod consists of a small group of AI builders focused on specific outcomes, often working across disciplines. For example, engineers could take on design work, depending on the task. Some Meta employees have already begun referring to themselves as AI builders on LinkedIn, Business Insider previously reported.

These pods are led by Pod Leads, who oversee day-to-day operations. They are, in turn, overseen by Org Leads, who also manage performance reviews and oversee promotions — processes that will be supported by unspecified "AI systems."

The memo said that the overall team size will remain the same under the new structure.

Meta laid off hundreds of staff on Wednesday, and this cut affected staff in Reality Labs, among other teams. A Meta spokesperson said the reorganization is not related to the cuts.

Have a tip? Contact Charles Rollet via email at crollet@businessinsider.com or on Signal and WhatsApp at 628-282-2811. Use a personal email address, a nonwork WiFi network, and a nonwork device; here's our guide to sharing information securely.

• Arm announced its own AI chip, the AGI CPU, and is partnering with OpenAI and Meta.
• The new Arm AGI CPU aims to address energy efficiency and memory constraints in AI data centers.
• Despite strong growth prospects, Arm faces competition from established players like Nvidia and AMD.

Arm has long run its business as an architect behind the scenes, designing chips that power almost all the world's smartphone and making money off royalties from the chips it designs for customers.

Now, Arm is changing it up by announcing its own AI chip, the Arm AGI CPU.

Arm CEO Rene Haas said Tuesday at a company conference that this massive pivot wasn't just an internal strategy shift—it was a direct plea from the world's most powerful AI giants. The company name-dropped OpenAI and Meta as major partners for this chip.

"The biggest reason we're doing this is that our partners have asked for it," Haas said Tuesday.

With energy constraints and memory shortages, the AI boom has created a massive bottleneck in data centers. Faced with this demand, Arm stepped up with an AI chip that it says is more energy-efficient. Arm says it sees a $1.5 trillion market opportunity as it moves into AI chips for cloud, edge, and physical AI.

Arm stock was up by more than 18% on Wednesday. Mizuho analysts wrote that they see "strong growth opportunities" for Arm in AI infrastructure and the automotive industry. Bank of America research analyst Vivek Arya wrote in a note to investors that the company's outlook could be "too ambitious."

Meta and OpenAI partner with Arm

Meta has been building out data centers at a massive scale to power its apps and its latest superintelligence ventures. Santosh Janardhan, head of infrastructure at Meta, said Tuesday onstage that its coming "Hyperion" cluster could draw 5 gigawatts of power, enough to power 50 towns the size of Palo Alto.

"If we met the performance, we couldn't get the power. If we got the power, we wouldn't get the performance," Janardhan said.

This sparked an engineering project within Meta, where engineers were "working 'round the clock" to port its systems to Arm in three months, said Paul Saab, a Meta engineer.

"I didn't even ask my boss here for permission to buy these machines or even start the project," Saab said onstage.

While Saab says he saw major performance benefits, at the time, there wasn't an Arm chip available to buy.

OpenAI faced a similar problem. Its compute demand has grown massively as it trains and runs its ChatGPT models, its AI coding tool Codex, and more.

"That is one of the most common things I hear inside OpenAI. I need more compute," Kevin Weil, vice president of OpenAI for science, said onstage, adding that it needed chips that were energy-efficient.

Arm said it expects this chip to generate $15 billion in revenue by fiscal 2031.

The chip market is 'getting very crowded'

Arm faces the risk that the CPU market is "getting very crowded," Arya wrote in his analyst note. Other competitors, such as AMD, Nvidia, and Intel, have more CPU products and more established customers. Notably, both Meta and OpenAI also work with AMD and Nvidia, which could leave "limited" opportunity for Arm's new CPU, Arya wrote.

"Moreover, the bigger AI grows, the more pressure ARM's smartphone/consumer markets would have from limited memory supplies," Arya wrote.

That said, the increasing demand has led many customers to turn to chip companies beyond Nvidia for their computing needs. Both Meta and OpenAI also work with Broadcom to build AI chips.

The rise of AI agents has also led to greater demand for inference, or how AI models draw conclusions and make predictions. While Nvidia's core AI chips, the GPUs, dominate AI training, CPUs like Arm's AGI CPU can also help with inference. Nvidia also recently made moves into this market.

Have a tip? Contact this reporter via email at rmchan@businessinsider.com, or Signal at rosal.13. Use a personal email address, a nonwork WiFi network, and a nonwork device; here's our guide to sharing information securely.

• Goodbye, Sora!
• I loved you at first — it was so much fun making silly videos of my friends.
• After a few days, I got bored and moved on. Apparently so did everyone else.

Goodnight, sweet Sora. You were a wonderful tool for trolling my friends, but you burned too bright (and used too much compute) to stay around in this harsh world.

For a brief moment, I absolutely loved Sora. I loved making silly videos of my friends and me. I loved that I could use my friend's face to put them in ridiculous situations, like falling over while roller-skating at their desk, experiencing gastric distress, or singing in a ska band.

I was addicted to making these, churning them out, often starting a new one while waiting for the previous one to render, and sometimes hitting the hourly limit that OpenAI had to impose after some people (oops) were burning through all that free compute. I drained my phone battery by midday.

But a few days after its initial launch, the small handful of my friends and colleagues who had any interest in joining had already joined. None of my normal friends who didn't work in tech or media had any interest in this at all and found it fairly unpleasant.

Also, I couldn't help notice that on the feed of videos, there seemed to be very few women using the app, or at least allowing others to make videos featuring their likenesses. That makes sense; women's experience on the internet has rightfully informed them that it would be a very bad idea to allow strangers to make videos with your face. What I discovered was that Sora had a pervert problem: Although nudity or sexual content was banned, people were making non-nude fetish content like feet videos with random women's likenesses.

Now, OpenAI announced on Tuesday that it will be shutting down Sora, its stand-alone video generation app, and its deal with Disney is dead. An OpenAI spokesperson told Business Insider that the company is focusing its resources on other parts of the business. It seems that Sora was one of the "side quests" that was a distraction and a drain on compute.

Sora became a bore-a

Eventually, my friends all seemed to get bored with the app. As I do at most parties, I stuck around longer than everyone else, but eventually I, too, found that the novelty had worn off. I rarely opened the app after the second week.

This was, I imagine, a problem: making videos of yourself is fun, but watching videos that strangers make of themselves is not fun. The idea of a social feed of AI-generated videos is simply not something that people are interested in. Around the same time, Meta also tried this with an app of AI videos, and it was even more boring.

The last few years have taught us that humans — myself included — have a nearly endless capacity to watch an algorithmic feed of vertical short-form videos. However, it seems clear that this only applies to human-made content: videos of people putting on makeup, dancing in their kitchens, lip-syncing, debating, whatever. A social feed of AI video simply doesn't work.

I am not sure that OpenAI was truly trying to create a successful social video feed; it seems more likely that this was a small experimental effort that ran its course and they're moving on.

I'll miss Sora in the way I miss something like ChatRoulette or BeReal: It was really fun for a short time, and then not at all, and I have zero desire to ever revisit.

Rest in peace, Sora — and thanks for the stupid memories.

• Nvidia CEO Jensen Huang urged all workers, from farmers to electricians, to embrace AI.
• He told podcaster Lex Fridman that the technology could elevate blue-collar jobs, such as carpentry.
• Blue-collar has generally been viewed as less likely to be affected by AI disruption than white-collar jobs.

Artificial intelligence isn't only coming for office jobs — Nvidia CEO Jensen Huang says blue-collar workers should be paying attention, too.

Huang leads one of the biggest chipmakers fueling the AI revolution. He joined Lex Fridman's podcast in an episode published Monday to discuss everything from AI in space to work.

While blue-collar jobs have been considered relatively safe from AI disruption compared to tech roles like engineering, Huang said workers in every profession, including farming and electrical work, should use artificial intelligence to help future-proof their jobs.

"If I were a farmer, I would absolutely use AI. If I were a pharmacist, I would use AI," Huang said. "I want to see what it could do to elevate my job so that I could be the innovator to revolutionize this industry myself."

For example, he said coding represents a big opportunity for carpenters, and he would go "completely berserk" using AI if he were in that line of work.

"A carpenter with AI is also an architect," he said. "They've just increased the value that they could deliver to the customer. Their artistry just elevated tremendously."

Huang has said before that he is "certain 100% of everybody's jobs will be changed" by artificial intelligence, and that while he expects some jobs to be lost, many will also be created.

Many tasks, for example, will be automated, and those jobs will be highly disrupted, he said on Fridman's podcast.

But, he said, "If your job's purpose includes you … then it's vital that you go learn how to use AI to automate those tasks."

Anxiety grows alongside AI

As AI advances, so has anxiety around job security. The fears aren't unfounded. Companies have slashed thousands of jobs in the name of prioritizing new technology and automation.

Huang's solution: Become an expert in AI, no matter what your job function is.

It could be the difference between landing a job and ending up unemployed. In almost every case, Huang said he'd rather hire the candidate who's an AI expert over one who isn't.

"Every college student should graduate and be an expert in AI," Huang said.

It could help them stay ahead of the curve as AI quickly advances.

The next phase of AI is already here

Artificial general intelligence is a form of AI that elicits anxiety or excitement among the field's most advanced minds. It's the idea that AI will one day meet or surpass human intelligence. Huang said that the age of AGI is already here.

Fridman asked if AI could do Huang's job of starting, growing, and running a successful tech company worth more than $1 billion.

It's possible, Huang said.

He also said, "It's not out of the question" that chatbots like Anthropic's Claude could design an app that billions of people would use for $0.50 apiece, and then go out of business shortly after, similar to websites that went bust in the dot-com era.

Even his job running one of the most successful tech companies today isn't immune to the effects of AI, he said, encouraging everyone to jump on the technology before they're left behind.

"Go see what it can do to transform your current job, elevate yourself," Huang said.

• San José airport starts testing an AI robot called José to assist travelers.
• The pilot test launched on Tuesday amid travel chaos at many US airports.
• Some TSA workers have stopped coming into work due to a government shutdown.

One of Silicon Valley's main airports just made its newest hire, a robot named "José."

San José Mineta International Airport is turning to artificial intelligence to ease the strain of modern air travel, debuting "José," a humanoid robot, as some US airports grapple with staffing shortages and widespread delays.

Developed by Silicon Valley startup IntBot, José is designed to greet passengers, answer questions, and provide real-time updates while autonomously navigating busy terminals.

The robot will be stationed in SJC's Terminal B as part of a four-month pilot, "singlehandedly running his own gate," according to an email previewing the test that referred to José as the airport's "newest hire."

Airport officials said the launch highlights San José's role as a testing ground for emerging technologies to improve customer service.

"By piloting IntBot, we're exploring how artificial intelligence can enhance the passenger journey while reinforcing SJC's role as the gateway to Silicon Valley," said SJC Director of Aviation Mookie Patel.

The timing is notable. Airports across the US have been hit by long security lines and travel chaos, driven in part by many Transportation Security Administration workers not reporting to work during a partial government shutdown. With TSA agents going unpaid at the height of the spring break season, some airports have struggled to maintain normal operations.

José the robot represents a broader push to automate parts of the airport experience, from passenger assistance to information delivery.

SJC officials said the pilot will help evaluate how multimodal AI, combining vision, audio, and language, performs in real-world environments.

The future of air travel may include a robotic helping hand — and it can't come fast enough for weary vacationers stuck in long lines.

Sign up for BI's Tech Memo newsletter here. Reach out to me via email at abarr@businessinsider.com.

• BNY's CEO, Robin Vince, is all in on AI's role in steering the bank's future.
• Now, some managers oversee the bank's 140 digital employees, a form of agentic AI.
• We spoke to Vince and a BNY managing director about the program.

Despite its 240-year pedigree, BNY isn't showing its age.

Under CEO Robin Vince, who took the reins in 2022, the firm — founded by Alexander Hamilton — is aggressively embracing AI. Recently, it has begun entrusting some managers with oversight of a contingent of new workers who don't even require a chair: the digital employee.

"All digital employees report to a human manager," Vince said in an interview with Business Insider this month in Palm Beach.

These digital employees create a layered effect with the company's agentic products, in which a single entity coordinates the activities of multiple individual agents. The digital workforce is more than 140 agents strong, each one with roughly two dozen skills, give or take, comprising their suite of abilities.

And, just like humans, they're held accountable for their work — with performance reviews.

After executing a variety of tasks humans might find tedious, the digital employee presents it to "the human who's responsible for the process — 'I've just done three quarters of the work for you. And by the way, I did it in 10 minutes instead of what would have otherwise been two weeks," the CEO explained.

About 100 managers across the firm oversee digital employees, including Rachel Lewis, a managing director and a two-decade BNY veteran who now serves as head of AI enablement for operations. Appointed to the role this year, Lewis is now helping teams across the bank build and deploy digital employees within their day-to-day workflows.

"We're kind of transferring the mundane to the machines," she said, describing how the tools are taking over routine processes and shifting how work gets done.

Lewis told Business Insider she works closely with teams across BNY to help them develop their own digital employees — often starting with ideas that come directly from the people doing the work and turning them into tools over time.

"The person that came up with that idea actually gets the opportunity to build that digital employee," she said. As teams begin to incorporate them into their workflows, she added, the technology starts to feel less like software and more like part of the team. "It's just almost having a virtual teammate as part of your group."

170,000 hours of training

To prepare for the AI age, BNY implemented a massive 170,000-hour AI training program for its 48,000 staffers. "Everyone in the company has done two to three hours," he said. The goal was to turn employees into a new class of supervisors who managed, rather than competed with, the machine. "We're investing in our people, because I want them to be the unlockers and users of AI," Vince added.

Last week, he sent a memo to several thousand of the firm's senior leaders pointing to some of the firm's past efforts in AI and encouraging them to be proactive in continuing to incorporate it. "We have an obligation to our company to capture this opportunity," Vince wrote in the email, whose subject line was "Reimagining BNY."

"This is a fundamental leadership shift, not simply a capability shift," he added. "It will require each of us to lean in and role-model how to engage with AI and how to harness it to solve problems."

Speaking to Business Insider, Vince described his first personal deep dive into AI as a "summer project" that kicked off in 2023 and never ended.

It was sparked by a YouTube video he saw that broke down the functionality of Tesla's Autopilot 12. He watched as the car observed human behavior and applied what it saw to navigating a stop sign, rather than adhering to a few rigid lines of code. "It was very clear to me that the future of AI was going to be learning to make decisions," Vince said. He wanted to bring that same adaptive intelligence to the bank. "It was highly applicable to our businesses," he added, "and it would be able to be a very fundamental input to how we actually ran the company."

Expanding the digital workforce

While some of the earliest digital employees have applications focused on straightforward fixes like data repair and data capture, Lewis said the tools that have stood out most are those that make it easier for employees to build and refine their own digital employees.

Building a digital employee starts with observing how work is actually done. Teams record themselves completing tasks step by step, allowing the system to analyze different approaches and identify the most efficient way to perform the work. That output is then used to generate the instructions that guide a digital employee, which are refined over time as teams train the system on new variations of the task.

Lewis said that as digital employees become embedded in workflows, teams are also treating them more like members of the workforce. "There is a performance review," she said.

Managers evaluate how the systems perform by reviewing outputs, identifying where they skip tasks or "didn't perform as expected," and feeding that work back into the system to be retrained on new variations and edge cases.

"We're continuously monitoring them," she added. "Every week it gets a little bit better."

Even as it expands its digital workforce, Vince said there are no plans to cut back on human capital; these tools, he said, are meant to supercharge their workflow, but not take responsibilities out of their hands. "I speak to CEOs who say, 'We're going to downsize, massively, our campus program.'" Vince's reply? "Why would you do that?"

"We've got the opportunity to have young people who are pre-trained in AI, enthusiastic, and be able to add to our business in different ways," he said.

• Srdjan Stakic vibe-coded a security camera system for his parents to ensure their safety.
• Stakic used vibe-coding platform Lovable to get started, as well as popular AI chatbots.
• His vibe-coded software became the basis for his AI-assisted startup Alvis.

This as-told-to essay is based on a conversation with Srdjan Stakic, 49, a former film producer who vibe-coded an AI system to monitor his elderly parents and detect falls. He's now launching a company that aims to offer the technology to others. This interview has been edited for length and clarity.

When I was diagnosed with stage four cancer two years ago, AI became essential.

Everything was happening so quickly: The doctors would talk to me for 15 minutes and leave me with more questions than answers. AI gave me an objective way to document and make sense of what was happening.

I'm now in remission. As my health improved, my parents' health declined, and I began helping them with cooking, cleaning, and medical appointments.

English was not their first language, and communicating with healthcare providers was tough. I recorded our conversations with their doctors and compared them to the after-visit summary using AI. I would put together all this information and translate it into Serbian for them.

But I soon wanted more than what the chatbots could offer. I wanted a system that could observe what was happening with my parents, or any other patient, and assess it through the lens of safety and dignity. I would think of how much guilt I'd feel if something happened and I wasn't there. What kind of son would I be?

I had never coded before, and I didn't have millions for an initial investment

I don't have a background in coding. I have a doctorate in health education and a master's in film production, and I have produced some films of my own.

I started outlining my idea with Gemini and ChatGPT to examine it from a tech and ethical standpoint. I built this document of what I wanted to achieve. I kept asking my family how they wanted to be treated in each scenario — like a fall or medical emergency — and I wanted to make the system flexible.

Then I transferred to Lovable. Lovable gave me a live development environment where I could describe what I wanted, see it built in real time, test it, and iterate. It connected the pieces, the frontend, the backend, the database, the authentication, the integrations, things I did not even know I needed until they were there. The chatbots helped me plan. Lovable helped me build.

I uploaded hundreds of training videos for nurses and healthcare providers to train the AI. I created a high-fidelity validation pipeline and a labeled dataset. I labeled real-world caregiving footage with established clinical benchmarks, like Stanford's C-I-CARE framework. When you approach a patient and introduce yourself, you tell them why they're there, you ask the patient's name and pronouns, and you introduce what you're about to do. You explain next steps and see if they have any questions or concerns.

I also started building an AI equipped with cameras to identify falls. I would fall in the middle of my living room and see whether the system recognized that and how long it would take.

It took me a few months to make it work

I tried different cameras and protocols, but ultimately, I had to hire an IT company to help me connect multiple cameras. The system can now identify a fall and send notifications to loved ones or EMS, and provide their location with a brief summary of their health records. The system also analyzes interactions between caregivers and my parents. It's sophisticated enough to analyze in real time — based on audio and video — if a caregiver is being rude or unprofessional. My parents have felt safer since I built this. I also built a feature that scans their environment for any trip hazards, such as cables.

I don't want to spy on my family, so I don't actively review all the video footage. When a concern is flagged, the system clips approximately 30 seconds around that moment and notifies mewith a summary of what it observed and why. It can also generate an advocacy letter from that same analysis: what was said, what was done, and how the interaction compares to the C-I-CARE framework to evaluate caregiver conduct.

I launched a company to offer this tech to others

This all started as an idea for my family, but the more I talk about it, the more people tell me they wish they had this for their parents. So I decided to launch a startup, called Alvis, to make this system available to others.

It detects falls in real time, recognizes when a caregiver goes above and beyond, and generates advocacy letters when something goes wrong. It's in private beta and accepting waitlist applications for our pilot cohort, launching April 13. The model will be a monthly subscription, similar to what families already pay for camera cloud storage, with a premium tier for AI-assisted analytics.

This week, my mom was hospitalized, and I used AI in four ways

First, I used it as a real-time medical interpreter: Every lab result went straight into Claude, so I understood what was happening immediately, not the next morning when a doctor was free.

AI was also my clinical advocate. When a history and physical exam understated her cancer history, Claude caught it. When her glucose started climbing from steroids, Claude flagged it.

Third, I used AI to translate updates into patient-friendly language in both English and Serbian.

Finally, Alvis — the camera system I designed — was running live in her hospital room all night, with her permission and a nod from her care team. It picked up her saying in Serbian, quietly, that she had endured too much. It flagged when I visited, and we recorded ourselves together.

It's amazing to see how vibe coding is democratizing access to AI tools. You can build a company that helps a very niche group that needs a specific thing. I still don't fully understand code or the extent of what I built, but it seems to be working.

• Mark Cuban said he bought a Mac Mini to fight a surge of AI-generated emails.
• He said he is training AI to auto-unsubscribe from spam flooding his inbox.
• Cuban believes AI outreach is a trial phase, and response rates will likely eventually drop.

Mark Cuban says the rise of AI-generated cold emails has gotten so overwhelming that he is now fighting back with AI of his own.

Speaking on the live-streamed tech show TBPN on Thursday, the billionaire investor said he recently bought a Mac Mini to help manage the growing flood of inbound messages.

"I do what everybody else does. I bought a Mac Mini," Cuban said.

Beyond AI-generated emails, he said the issue is unwanted email subscriptions.

"It's not even like the cold emails because that's pretty obvious," Cuban said. "It's people subscribing me to shit."

His fix, he said, is to use AI to automate the cleanup.

Cuban said he is training systems to take advantage of Gmail's built-in unsubscribe button, effectively creating a loop where AI filters out AI-generated noise.

"You just got to train it to hit the unsubscribe button," he said. "Then, I just review it and all that shit, so it's still a work in progress, but at least I have a path."

Cuban didn't immediately respond to Business Insider's request for comments.

A 'trial and error phase'

The approach reflects a broader shift in how executives are using AI to manage their inboxes.

LinkedIn CEO Ryan Roslansky has said he uses Microsoft's Copilot for "almost every" high-stakes message, and executives across industries, from tech to retail, recently told Business Insider's Ana Altchek that they rely on AI for day-to-day communications and reviewing documents.

Cuban framed the current moment as a trial-and-error phase, where people are testing what works and what doesn't.

"We're in that trial and error phase where people are like, 'We're going to try it, see what happens,'" he said, adding that response rates will likely fall as more AI-generated messages flood inboxes.

"Then they'll get bored, and then it'll drop off," he added.

• Meta might not yet have the best AI models, but it could be winning on another crucial front.
• Meta is considering major layoffs across the company, Business Insider reported.
• A top analyst suggested these cuts could signal that Meta's AI transformation is underway.

Could more Meta layoffs mean the company still has bloat to trim — or could it signal its AI investments are actually starting to pay off?

A note from a top Wall Street analyst said on Monday that any head count cutting from Meta could actually be a sign that it's successfully rebuilding itself as an "AI-forward" company.

That could be bad for its rivals.

While Meta's deep investment in AI has so far not produced leading models like Google's and OpenAI's, Bernstein's Mark Shmulik said Meta's aggressive push to overhaul itself into a top-to-bottom AI company could put it ahead of competitors and trigger a "wave of panic" as competitors scramble to copy it.

Meta is pouring hundreds of billions of dollars into building out AI data centers and luring talent to shore up its AI research teams. Last week, Business Insider reported the company was weighing up layoffs, with some managers being asked to draw up cost-cutting plans.

Bernstein's Shmulik said this could be a signal that Meta is winning on a crucial front in the AI wars. While companies can win with world-class frontier models, they can also beat the competition by deploying AI so deeply across the core business that their competitive moat "widens beyond dispute," Shmulik wrote.

"Meta has already demonstrated the compelling returns they're seeing from deploying AI to core workloads," wrote Shmulik. "But if the company can now re-design their operations from the ground up to be AI-forward, their potential cost and performance advantage could be insurmountable."

By one metric, Zuckerberg's efficiency drive over the past three years has paid off. Revenue per employee has steadily increased over that time period, with the company overtaking Amazon last year, according to data shared in the Bernstein note this week. Pinterest was the only company with a higher ratio.

At the same time, Meta's capex and R&D spend per employee have significantly outpaced rivals, according to the Bernstein report, which could point to a reason for the potential layoffs.

Investors appeared to react positively to Meta weighing up further cuts, with the company's shares up about 2% early Monday.

AI-washing? Maybe not.

Like other Big Tech companies, Meta has moved quickly to chase AI.

It has also been aggressively driving AI adoption internally. The company said it would start grading employees on their "AI-driven impact" in performance reviews starting this year, and has tracked how some teams have been using the tools, Business Insider previously reported.

Companies including Atlassian and Block have cited AI as a reason for recent layoffs, raising the question of whether some leaders are "AI-washing" and using the technology to camouflage other reasons for cuts, such as financial problems or overhiring during the COVID pandemic.

Bernstein's Shmulik said that while AI-washing was possible in Meta and other companies' cases, he said that layoffs could now be seen as evidence that the company is seeing efficiency gains.

The company eliminated more than 20,000 jobs in late 2022 and early 2023 as Zuckerberg declared a "year of efficiency," cutting non-technical roles, flattening management layers, and lifting what had been a sagging share price.

If Meta repeats a similar cycle for the AI era, it could set the mold for what a truly AI-first company could look like, Shmulik said.

"If one major player is able to redraw the blueprint for an AI-enabled organization, others will rush to replicate it... and we wonder if this could trigger a cascade of hurried pivots, half-formed strategies, and reactive restructuring across the ecosystem," he wrote.

Have something to share? Contact this reporter via email at hlangley@businessinsider.com or Signal at 628-228-1836. Use a personal email address and a non-work device; here's our guide to sharing information securely.

• Saul Van Beurden thinks employers and employees share responsibility for AI adoption.
• Wells Fargo doesn't mandate AI use; instead, it aims to generate "grassroots enthusiasm."
• Van Beurden said employees need new skills to stay competitive for both redeployment and new jobs.

Saul Van Beurden is the man helping Wells Fargo confront a question hanging over banks of every size: What happens to jobs in the age of AI?

He and his central team can't, and shouldn't, figure out what an AI-ready Wells Fargo looks like alone. The bank must teach employees skills to stay competitive in a changing industry, and they must choose to learn them, Van Beurden said.

"You cannot deny things," Van Beurden, who is the head of AI and the co-CEO of consumer banking and lending, told Business Insider. "But how do you make it a thing where everybody has a role to play and takes their own accountability and responsibility?"

The bank is leaning on AI literacy programs and demos, among other things, to hopefully inspire "grassroots enthusiasm." The goal is to make employees comfortable enough with the technology that they can be redeployed if their jobs change, or competitive in the job market if they leave Wells Fargo, he said. Wells Fargo doesn't mandate AI usage, even as it bets the technology will help supercharge its growth following the Federal Reserve's decision to lift a $1.95 trillion asset cap.

Van Beurden thinks that fluency starts outside the office. He's trying to build an agent to help pull documents for his 2026 tax returns, and believes it's crucial for employees to use AI in their personal lives, too.

"It's really important to have that personal usage, to understand the power of what it can do. And then we are enabling that and allowing that to happen at the workplace," he said.

Still, Van Beurden emphasized that everyone needs to "stay cognitive," since AI could generate all of our ideas if we let it. He suspects that most college students are comfortable with technology but should invest time in activities like reading or playing chess. Staying sharp, he thinks, will help them in what's broadly a brutal job market.

Wells' workforce, like many of its competitors, is already changing because of AI. The bank's CEO, Charlie Scharf, said in November that it will probably "have less head count as we look forward," and added in December that generative AI has already made engineers up to 35% more productive.

Van Beurden didn't say whether the bank would need 30% fewer engineers as a result or whether it would necessarily alter hiring, leaving it at, "it's a great question." Instead, he said that growth and head count aren't always one-to-one.

"How great is it to grow without the need to hire people, because you have created the capacity to take on more clients, to take on more customers with the same amount of people?" he said, calling AI the "ideal tool" for that growth. Wells Fargo recorded $21.3 billion in revenue in the fourth quarter, up 4% year over year; revenue in its consumer bank, which Van Beurden oversees, rose 7% year over year.

The leaders of other big banks have also said that AI will likely eliminate some jobs and slow hiring, both publicly and in internal memos. JPMorgan CEO Jamie Dimon has said his bank has "huge redeployment plans."

Efficiency promises and big technology budgets aside, the head count cuts haven't yet materialized at most banks. Around 60% of 240 financial services CEOs surveyed by EY said they expect AI investments to maintain or boost their head count this year.

• A version of this story originally appeared in the BI Tech Memo newsletter.
• Sign up for the weekly BI Tech Memo newsletter here.

A few Fridays ago, I was feeling smug. I'd just sent another Tech Memo edition telling subscribers to stop worrying about AI eating tech jobs because Anthropic, the leading AI company pushing this narrative, is hiring so many engineers.

So clever! Until I got an email from a reader, Kiran Maya Sheikh. She has a computer science degree from the University of California, Irvine. It's a great school, and she graduated with an impressive GPA. And yet, she's struggling to land that all-important first full-time software engineering job.

"It's bad advice to 'not worry,'" she wrote. "AI is causing disruption in this job market. Employers are prioritizing hiring experienced workers, but not new graduates."

This week, I interviewed Kiran for Tech Memo. It was an eye-opening view into the realities of the new AI economy. Here are the highlights from our chat, edited for clarity and length.

Alistair: What did you think you were signing up for when you first chose computer science as a degree?

Kiran: After getting into UC Irvine in 2020, I took my first coding class and I really enjoyed it. The prospects at this time were that people were going into this major to get great jobs and it was very rewarding and I ended up liking the work.

What did you believe a career in computer science would give you financially, socially, and emotionally?

The dream at the time was definitely everyone was saying, "Let's go work for Google and the FAANG companies and get a six-figure salary." My motivation was just getting a stable job, getting enough money to take care of my family — what everyone wants. I expected that computer science would put me in a position to grow as a software engineer, first and foremost, and then maybe take me to more of the strategic side, the management side. The main thing that I did figure out was that I wanted financial stability and maybe financial independence as well.

Fast forward to late 2022, when ChatGPT launched. Did you see that as a tool at the time or a threat?

I was a hater at the beginning. Then, friends of mine started using ChatGPT and they're like, "Oh, you can just use it like Google. You can just text it and it'll give you the answer." And honestly, my first thought was like, "That's a bit lazy. You can get more learning out of doing the work yourself." But the more time went on, the more that people were using it, and they started using it for class. Suddenly, I was ahead in class. I was doing the assignments well and understanding more.

Was there a moment when you thought generative AI might reduce the need for junior engineers, or do you even believe that?

We all know the current job market. It's not too hot and a lot of companies are citing AI as part of the reason for layoffs — but maybe that they were going to cut those jobs anyway. At the time though, while I was in school and using ChatGPT, I honestly didn't think it would get this far. I expected AI would be integrated into software engineers' work and companies would start integrating it, but I didn't realize there would be potential for it to take over jobs that I was looking for.

I don't think I was very attentive to the job market situation at the time, and I wasn't really thinking that far ahead. More of my worries at the time were just getting that first entry-level position. And I just thought it would be simple: I just get my degree and I would find a company that's hiring. Looking back, it was my mistake to not really research the current job market and maybe what some people were predicting about AI.

I didn't see it coming either. Few people did. Anyway, describe the moment when you realized the job market had changed?

I was already graduating, so this was after June 2025. I was getting into the reality of having to find my first job, and that's when I definitely started noticing something was wrong. A lot of my classmates, I haven't really heard of them getting any opportunities. Everyone's submitting so many resumes and there's a race to use AI to enhance resumes and send them out as fast as you can. And it seemed a lot more intense than I was prepared for.

A lot of my classmates and even students I know who are still in school are not even landing internships right now. It's not looking great. It's a very tough battle right now. So many people are quitting or getting fired or pivoting and there's new grads. Everyone is bracing, and it's a bloodbath right now.

Do you feel like you're competing against AI or laid-off senior engineers or both, or something else?

My fight is definitely with AI and all the competition with entry-level graduates — especially because AI is known to take over more junior roles. So it's important that we stay more relevant and offer something that AI can't. Scrolling through LinkedIn and on my job portals, I see more offers for mid-level positions, but I don't see as many for entry-level roles. So it's like I'm fighting AI and all these other graduates for roles that don't exist yet.

This job search so far, what has it done to your confidence?

I try to be optimistic. I am lucky to have a better situation than some other people do. I'm living at home with family, so I don't have to worry as much about expenses. Still, if I weren't doing anything about my situation, I would feel pretty bummed. I'd feel kind of trapped.

But I've been trying to work on building my network, finding people I know and learning from other people, just finding communities to be involved with. That's really helped my confidence because I find professionals that are trying to help — they are aware of the job market and they know how hard it is to get that first job. The one saving grace in this tough situation is definitely the community I've found and the people I know who are helping me through it.

Did you ever question your decision to study computer science?

Yes, I did question it. But I remember that I do like computer science and I did like what I learned. I really enjoyed my classes and programming. And instead of turning to a new discipline, I think I prefer to just specialize and find out new information and stay ahead of the news. And like I said, offer something that AI can't.

Do you feel like you were trained for a version of the tech industry that no longer exists?

I am a little salty, about this, if that's the right word. During my time at school, a lot of what the degree was about was learning the basics of software engineering. You learn programming languages and you learn how to set up your development and deployment. But right now there are so many more tools and I think that's the constant thing with the software engineering and the tech industry. There's always new technology and there's a lot of learning you have to keep up with.

But with AI in particular, I felt like I graduated a bit too early. Because now AI will probably be more integrated into learning. I had so many professors that were more welcoming towards AI. I remember a really cool professor who shared a website that would let you make your own LLM. And it's really useful stuff, but it wasn't part of the curriculum. It will be now, but I won't be there to see that change.

What I'm doing to help with that, and make the amends, is volunteering and doing more work on the side that involves newer technologies to just stay fresh and relevant and use all these new AI tools and see how I can leverage it.

If a high school senior asked you today whether they should major in computer science, what would you tell them?

It depends on what interests them about computer science. If it's absolutely something they're interested, they love learning about the technology and they want to code, I would still say go for it, but I would recommend how to position yourself for after college.

You need to start much earlier now, networking and knowing how to speak with people and how to apply, how to write a resume. And those all are also much more important now at the start of college, especially getting internships, if at all possible.

So, I would definitely recommend studying computer science, but being realistic about the opportunities available and keeping up with the news and the job market.

What would you say to potential employers out there?

The focus should still be in hiring entry-level talent if possible. I know it's tough with the current market and the economy and what's going on in the world right now. But entry-level talent is still important because you need to build this generation of professionals so that the future will have people to rely on. AI is still uncertain right now. People are still figuring out how it is impactful and it doesn't help to just force it upon your company.

Sign up for BI's Tech Memo newsletter here. Reach out to me via email at abarr@businessinsider.com.

• Anthropic's Jeff Bleich says AI will end the billable hour's dominance in legal billing.
• Billable hours mean lawyers get paid more when they spend more time on work.
• But AI tools eliminate "tedious" work, which devalues the time lawyers spend overall, Bleich said.

The billable hour's time is approaching midnight, according to Anthropic's top lawyer.

"I don't think the billable hour is the solution, and we've known it for a long time," Jeff Bleich, the AI company's general counsel, said Thursday.

Speaking at the American Bar Association's White Collar Crime Institute in San Diego, Bleich said that artificial intelligence tools are eliminating the need for companies to hire armies of lawyers to do lucrative yet "tedious" work.

"Now we've got a technology that's going to eliminate the sorts of things that allow people to become wealthy off of tedious work," Bleich said on the panel, alongside top lawyers at Google, IBM, and Liberty Mutual. "That was not what lawyers are trained to do, and not what we ultimately look to lawyers for."

The much-maligned billable hour is the standard method that law firms use to bill their clients.

Attorneys track the work done for each client, often in six-minute increments, tally them up, and charge their clients accordingly.

While the billable hour has been useful to help companies and other clients understand what they are paying lawyers for, it has also "created a wedge," Bleich said.

Under the current system, "the interests of firms are at odds with the interests of their clients," he said. Companies want lawyers to resolve problems quickly, but law firms get paid more when the work takes longer.

"Clients want you to solve the problem as efficiently as possible and with as little drama as possible," Bleich said. "And if you're a company, the bigger the case gets, and the more dramatic it gets, and the more complicated it gets, and the more work that has to be done — the more lucrative it is."

The other panelists largely agreed with Bleich's remarks.

"The value is no longer you putting in time," said Damon Hart, the top lawyer at Liberty Mutual. "The value is your strategy, your results."

Anne Robinson, IBM's general counsel, told the audience that she's open to working with them to figure out more creative billing methods.

"I'm open to firms coming and saying, 'I'd really like to work with you on this matter or this type of work, I get that the billable hour model is not one of aligned incentives, and so let's sit down and talk about what you expect as far as outcomes and how we can both get there in a way that reflects your pressures and your priorities,'" Robinson said.

Bleich said he still values the work of outside law firms, but wants them to find an alternative to the billable hour that works for everyone.

"We're not going to sort of cheap out and starve you," Bleich said. "On the other hand, you have to have an economic model that works. And the firms that adapt to that faster and better will be leapfrogging other firms, because they'll be more attractive to work with."

Bleich's comments come at a critical moment for Anthropic, which sued federal agencies this week after the Trump administration effectively blacklisted it following the collapse of contract negotiations with the Department of Defense.

In the lawsuit, Anthropic is represented by WilmerHale, one of the law firms that Trump targeted last year with an executive order that was quickly blocked by a federal judge.

"I like firms that show some spine," Bleich said following the panel, when asked about using law firms that fought back against Trump's executive orders targeting them. He declined to comment on the lawsuit itself.

WilmerHale is distinguished in another way: Reginald Heber Smith, who in the early 20th century managed the Big Law firm — then called Hale and Dorr — is widely credited with inventing the billable hour.

• Sam Altman said AI would in the future be sold like electricity and water, metered by usage.
• "We see a future where intelligence is a utility," the OpenAI CEO said.
• Altman suggested demand is surging and compute will decide who gets access.

In the future, you could have another utility bill to pay for: artificial intelligence.

That's according to Sam Altman, who says AI will eventually be bought and sold as a basic utility like electricity and water that's metered by usage.

Speaking at the BlackRock Infrastructure Summit in Washington, DC, on Wednesday, the OpenAI CEO said tech companies like his are building toward a future where intelligence is delivered on demand.

"Fundamentally our business and I think the business of every other model provider is going to look like selling tokens," Altman said, referring to the units AI systems use to process and price input and output data.

"We see a future where intelligence is a utility like electricity or water and people buy it from us on a meter and use it for whatever they want to use it for," he added.

In that world, compute capacity determines who gets access — and demand for AI is only going up, Altman said. Compute capacity is the processing power required to train and run AI models, determined by infrastructure such as chips and data centers.

If OpenAI doesn't build enough compute capacity to meet demand, Altman said, it either "can't sell it or the price gets really high." That would push AI access toward the wealthy, or force governments to decide how limited compute should be distributed, he said.

The infrastructure sprint

Major tech companies are set to spend hundreds of billions of dollars this year on compute to meet soaring demand for AI.

In her keynote at CES 2026 in January, AMD CEO Lisa Su said the world will need more than "10 yottaflops" of compute — a scale 10,000 times larger than global AI capacity in 2022 — over the next five years to keep up with growth.

Powering that expansion is a significant infrastructure challenge.

AI data centers can consume as much electricity as small cities, and the strain on the US power grid — along with transformer shortages and slow permitting for transmission lines — could become a bottleneck.

In an episode of the "Moonshots with Peter Diamandis" podcast in January, Elon Musk said that electricity generation is now the limiting factor in scaling AI, predicting China could outpace the US in total AI compute because of its faster energy build-out.

Inside tech companies, compute is a valuable but sometimes scarce resource. Engineers are competing for access to GPUs, and some job candidates now ask about their AI compute budget alongside salary and equity.

Last December, OpenAI President Greg Brockman said the company, which has committed roughly $1.4 trillion on data center projects over the next eight years, wants "to be ahead of the curve," but said, "I don't think we will be, no matter how ambitious we can dream of being right now."

At the BlackRock Infrastructure Summit, Altman said the goal is to move away from a world of being "capacity constrained."