Good morning. Everyone’s talking about Jensen Huang’s latest comments on whether AI will wipe out the workforce.
Such predictions are “hurtful,” the Nvidia chief said in an interview published Friday, and made by CEOs with “a God complex.” He added: “Before you know it, you know everything.”
Huang’s remarks, which included estimates about how many jobs AI has created, were seemingly meant to push back on critical comments made by Anthropic CEO Dario Amodei last year predicting an AI-driven bloodbath for white-collar jobs.
Make no mistake: Amid this AI arms race, the battle for mindshare among AI CEOs is as interesting as the battle for actual business. OpenAI chief Sam Altman adopted a bluntly patriotic stance. Google DeepMind’s Demis Hassabis pushed back on widespread AI-driven job loss. Anthropic’s Amodei predicted the loss of half of entry-level white-collar jobs within five years; Microsoft’s Mustafa Suleyman gave it 18 months. And xAI’s Elon Musk, ever subtle, predicted the end of all human jobs and called for “universal high income.”
My 2c? If any of them are right, it’ll be by sheer luck. But if we must “ground ourselves to talking about the facts,” per Huang, there’s one thing that’s certain: All have a vested financial interest in the growth of artificial intelligence.
Today’s tech news below. —Andrew Nusca
Want to send thoughts or suggestions to Fortune Tech? Drop a line here.
GameStop makes an unsolicited $56 billion offer for eBay
If you had this one on your bingo card, please tell me who’s going to win the next Super Bowl, would you?
GameStop, the popular video games retailer, is proposing to buy eBay for about $56 billion in cash and stock.
GameStop, it must be said, is worth about $12 billion on the New York Stock Exchange. Meanwhile, eBay is worth about $46 billion on the Nasdaq.
Huh?
GameStop has a roughly 5% stake in eBay already. It also has about $9 billion in cash. The retailer says it secured $20 billion of debt financing from TD Bank to assist in the deal.
“eBay should be worth—and will be worth—a lot more money,” GameStop CEO Ryan Cohen told the Wall Street Journal. “I’m thinking about turning eBay into something worth hundreds of billions of dollars.”
If Cohen’s name sounds familiar, it should be. The 40-year-old cofounded Chewy, the Amazon-for-pets site, in 2011. (He stepped down as its CEO in 2018.)
Cohen told the paper that he’s ready for a proxy fight and has no qualms about taking the offer to shareholders. He’s certainly motivated: Per his current compensation package, he’ll earn options on more than 171 million shares if he manages to elevate GameStop’s market cap to $100 billion…or about double the value of Ford Motor Co. —AN
Anthropic wants to sell AI tools to PE-backed companies
Anthropic is reportedly finalizing a deal to create a new joint venture with a handful of top Wall Street firms.
In the mix, according to the Wall Street Journal: Blackstone, Goldman Sachs, and the private equity firm Hellman & Friedman. General Atlantic and others are also reportedly involved.
The goal of the $1.5 billion venture, it seems, is to sell AI tools to PE-backed companies. The JV will act “as a consulting arm for Anthropic,” per the Journal, and “teach businesses how to incorporate AI across their operations.”
Anthropic’s bitter rival, OpenAI, is also said to be looking at a similar arrangement with PE firms.
It’s hardly an unprecedented move. Professional services have long been a natural extension for tech providers. Look no further than stalwart IBM, which earned $21 billion in revenue last year for its consulting arm and expects continued growth thanks to the AI boom; Microsoft, Hewlett Packard Enterprise, and Google Cloud offer similar in-house services.
What is unprecedented: How quickly Anthropic and its rival are entering the category. It took IBM 80 years to launch a consulting interest; Anthropic is on track to do it in five. —AN
Maryland bans ‘surveillance pricing’ in grocery stores
Is it fair to change the price of a product for sale based on a person’s location, demographics, or even search history?
No, according to Maryland Gov. Wes Moore. His state has become the first in the U.S. to ban the practice, which is also called “dynamic pricing,” in supermarkets and grocery stores.
“At a time when technology can predict what we need, when we need it, when we’ll pay for it, and also when we’ll pay more for it, and at a time when we’re watching how big companies are then using those analytics against us to make record profits…we are going to protect our people,” Moore said.
The state’s Protection from Predatory Pricing Act (HB 895), which takes effect Oct. 1, isn’t ironclad. Grocery focus aside, it also includes exemptions for loyalty programs and promotional offers. And it doesn’t stop a grocer from raising prices to everyone, then offering individualized discounts.
Still, it’s the first piece of legislation in what will likely be a wave in the coming months in the U.S. The states of Colorado, California, Massachusetts, Illinois, and New Jersey are considering restrictions on the practice in some form.
Don’t expect it to stop at groceries, either. The Federal Trade Commission (FTC) has previously documented examples of surveillance pricing at apparel, beauty, home goods, and hardware stores. —AN
More tech
—RIP Ask.com. The IAC-owned search engine once known as Ask Jeeves launched in 1997.
—SoftBank’s battery plans: Lithium- and cobalt-free versions for data centers as Japan works to reduce its reliance on Chinese metals.
—Amadeus will acquire Idemia for €1.2B. The former is the world's largest travel booking system, based in Spain; the latter is a French biometrics company.
—Linkerbot reportedly fundraises at a $3 billion valuation. The Chinese company is estimated to control about 80% of the global dexterous robotic hands market.
—90% of Nvidia’s production costs are attached to Asian suppliers, according to a new analysis, up from 65% last year. Why? The move to “physical AI.”
—Nintendo shares are down 45% since last August. Why? The global memory chip shortage is eroding profit margins for its Switch 2.
—OpenAI's o1 out-diagnoses triage docs. An AI accuracy rate of 67% in the ER versus 55% for humans, according to a Harvard study.
