Amodei and Altman Walk Back AI Job Fears as IPOs Near

Dario Amodei spent most of 2025 telling anyone who would listen that artificial intelligence could wipe out 50% of white-collar jobs within a few years. Sam Altman told his brother on the Uncapped podcast that "a lot of jobs will go away." Both predictions are now being walked back, at roughly the same moment, and with IPOs on both companies' horizons.

Changing the Calculation

Altman's reversal came at a Commonwealth Bank of Australia event in late May 2026. Speaking with the bank's CEO, Matt Comyn, Altman said he was "pretty wrong" about the economic disruption AI would cause. He explained that he had tried delegating his own Slack messages and email responses to AI, then found himself reverting to manual replies. "We really do care about our interactions with people," he said. "This thing is not something that I can imagine myself outsourcing to an AI anytime soon. It really updated me to thinking that the jobs picture is likely to be very different than we thought." He called the current outcome one he was "delighted" to be wrong about.

Amodei's shift was softer but pointed in the same direction. Earlier in May, he reframed the automation question using the Jevons paradox, the economic observation that more efficient use of a resource tends to increase overall consumption rather than reduce it. Rather than destroying jobs, he argued, AI would likely multiply output per worker, expanding what people do rather than eliminating them from the picture. That framing drew attention at the time as a meaningful rhetorical pivot from his earlier, starker language.

What the Data Actually Shows

The empirical picture is genuinely mixed. The Yale Budget Lab has found no statistically significant changes in occupational mix or unemployment duration in high-AI-exposure jobs since ChatGPT launched at the end of 2022. That is a notable finding, though researchers caution that the most disruptive effects of any broad technology shift tend to show up with a lag. U.S. tech layoffs through May 2026 have passed 115,000, already approaching the 124,000 logged across all of 2025, and companies including Meta, Amazon, and Snap have explicitly cited AI as a factor in headcount reductions. Whether that counts as automation displacing jobs or normal workforce rebalancing is something economists continue to argue about.

The Ramp AI Index, which tracks business software purchasing, offers a different angle. Anthropic's share of business AI spending reached 34.4% in April 2026, overtaking OpenAI's 32.3% for the first time. That enterprise momentum suggests companies are buying more AI tools, not fewer, which aligns with the output-multiplier framing Amodei has adopted. But buying more tools and reducing headcount are not mutually exclusive, and the Ramp data says nothing about employment outcomes.

"I thought there would have been more impact on entry-level white-collar jobs being eliminated by now than has actually happened." Sam Altman, Commonwealth Bank of Australia event, May 2026

The IPO Factor

The reversals are arriving at a specific moment. Anthropic is targeting an IPO as soon as October 2026, with Goldman Sachs and JPMorgan reportedly in early underwriting discussions. OpenAI is also preparing for a public offering, and SpaceX remains on a similar track. All three companies are navigating a public conversation in which fear of AI job displacement is one of the loudest themes.

The connection between softer job-loss rhetoric and upcoming stock listings is not subtle, and observers have not ignored it. A company going public with a narrative that it could eliminate half the white-collar workforce faces a different reception from institutional investors, pension funds, and regulators than one promising to make workers more productive. The framing shift is commercially convenient, which does not necessarily make it wrong, but it does make the timing worth noting.

Altman himself acknowledged that his reversal came partly from firsthand experience with the tool. That kind of personal data point is hard to dismiss. Amodei, for his part, has not abandoned concern about labor-market disruption entirely, only recalibrated the mechanism. As Anthropic moves closer to a public offering, how its founders talk about the technology's societal effects will carry more weight than any individual interview or podcast appearance. That weight is now shifting.