What's up: How AI small models do the real work; AI teaches the next generation of M.B.A.s; Palantir thinks college might be a waste.

Good morning. Does anyone even remember Halloween? There was so much happening this weekend, from the epic World Series to the New York Marathon, that Friday seems like ages ago. But here’s one aspect of Halloween that stuck in my mind: Friday’s edition of the BG2 podcast in which Altimeter Capital founder and CEO Brad Gerstner interviewed OpenAI CEO Sam Altman and Microsoft CEO Satya Nadella on many facets of the AI boom.

The episode delves into the relationship between Microsoft and OpenAI, the outlook for OpenAI’s growth, the evolving frontier of model development, the economic impact of AI, and much more. Here are a few highlights:

Altman on OpenAI revenue. Gerstner asked Altman how OpenAI, a company with about $13.5 billion in reported revenue, can make $1.4 trillion in spending commitments. Gerstner called it the single biggest question hanging over the market.

“We’re doing well more revenue than that. Second of all Brad, if you want to sell your shares, I will find you a buyer …. We do plan for revenue to grow very steeply,” Altman said. “We carefully plan, we understand where the technology, where the capability is going to grow, go, and how the products we can build around that and the revenue we can generate. We might screw it up.

“Like, this is the bet that we're making, and we're taking a risk along with that. A certain risk is if we don't have the compute, we will not be able to generate the revenue or make the models at this kind of scale.”

The $100 billion question and the prospect of an IPO. “But it does seem to me if you guys were, you know, are doing in excess of $100 billion of revenue in ‘28 or ‘29, that you at least would be in a position,” Gerstner said.

“How about 27?” Altman responded.

Nadella on the future of work. “There's a new way to even learn, right? Which is … how to work with agents, right? … When Word, Excel, PowerPoint all showed up in office … we learned how to re-think … Similarly, right now, any planning, any execution starts with AI, you research with AI, you think with AI, you share with your colleagues and what have you. So there's a new artifact being created and a new workflow being created.

“And that is the rate or the pace of change of the business process that matches the capability of AI. That's where the productivity efficiencies come. And so organizations that can master that are gonna be the biggest beneficiaries.”

AI has turned the M.B.A. case method on its head.

Some business schools are adapting. At Northwestern’s Kellogg School of Management, students no longer read through every available factoid on a real-life corporate challenges and write a memo. Instead, M.B.A.s must draw out details and data through open-ended conversations with AI chatbots and then craft a strategy, WSJ reports. 

Students who’ve worked through AI-guided cases say it more closely resembles what consultants and business leaders do in the real world.

“We started to fight fire with fire,” says Kellogg associate professor Sébastien Martin, who created the tool—not as a shortcut but to slow students down and force them to think and engage.

As AI enters the workplace, its human minders are not far behind.

AI startups like OpenAI, Anthropic and Cohere are increasingly recruiting and deploying forward-deployed engineers tasked with  working onsite with customers to help them customize their models, FT reports. 

This company thinks college might be a waste.

Data-analytics company Palantir is recruiting high school grads for its “Meritocracy Fellowship”—an experiment launched under Palantir CEO Alex Karp’s thesis that existing American universities are no longer reliable for training good workers, WSJ reports. The fellowship includes a four-weeks of humanities seminars on company-approved topics. Karp studied philosophy at Haverford College and holds a law degree from Stanford University.

Powerful frontier and reasoning models may grab the headlines, but that it is is the "small language models," that get the real work done at companies. 

Meet the AI factory. Where AI-powered companies have found success is by building their software and services more like an assembly line of AI, WSJ Tech Columnist Christopher Mims reports

In a factory, widgets travel down a conveyor belt, and workers tweak those widgets along the way. In what you might call an AI knowledge factory, chunks of data flow through pipelines of conventional software and are handed from one simple-minded AI to the next, with each one altering, sorting or transforming it.

As a group of researchers from Nvidia and the Georgia Institute of Technology wrote in a recent paper, the rise of AI agents (like our assembly line workers) is “ushering in a mass of applications in which language models perform a small number of specialized tasks repetitively and with little variation.”

Google is returning to Europe's debt market for the second time this year with plans to sell at least €3 billion to support its AI and cloud infrastructure, Bloomberg reports. Last week Meta Platforms, facing similar pressure to “aggressively” increase spending on AI, finalized pricing on a $30 billion six-part bond sale.

Standard Chartered CEO Bill Winters said this week that he sees a future in which nearly all global transactions are conducted on a digital blockchain ledger, CNBC reports.

IREN, a bitcoin mining and data center company, has signed a multi-year deal with Microsoft to provide access to Nvidia GB300 GPUs over a five-year term, WSJ reports. 

Kimberly-Clark has agreed to buy Kenvue for more than $40 billion, combining the maker of Huggies diapers with the owner of Tylenol and other consumer-healthcare products. (WSJ)

Democratic senators again urged President Trump to get involved directly in talks to end the government shutdown as the impasse entered a crucial week, with the lapse set to become the longest ever while pain for American households and travelers is deepening. (WSJ)

Democrats are favored Tuesday to win back the Virginia governor’s mansion, hold on to the New Jersey governorship and persuade California voters to back a ballot initiative that could give their party a fighting chance in the congressional redistricting arms race. (WSJ)

A Food and Drug Administration official who resigned on Sunday was sued by a Canadian pharmaceutical company, which accused him of soliciting a bribe and tanking its stock with false statements. (WSJ)

The WSJ CIO Journal Team is Steven Rosenbush, Isabelle Bousquette and Belle Lin.

The editor, Tom Loftus, can be reached at thomas.loftus@wsj.com.