Two of the most influential CEOs in tech spent 2025 warning that AI would gut white-collar employment. Now they're admitting they were wrong—and the timing couldn't be more interesting. OpenAI CEO Sam Altman and Anthropic CEO Dario Amodei have both publicly walked back their apocalyptic predictions about AI's impact on jobs, with both companies reportedly preparing IPOs valued at $1 trillion each.

Altman's Mea Culpa

In an interview with Commonwealth Bank of Australia CEO Matt Comyn on Tuesday, Altman said he was 'pretty wrong' about AI's economic impact—a stark reversal from his June 2025 warnings that entry-level roles faced serious risk. The OpenAI chief tried delegating his own Slack and email responses to AI as a personal experiment, only to find himself switching back to manual replies. 'We really do care about our interactions with people,' he admitted. 'This thing is not something that I can imagine myself outsourcing to an AI anytime soon.' Altman acknowledged taking 'a lot of flack' for his hype but defended his earlier stance as prudent caution: 'At the time I was like, "I see this is a real risk we should probably talk about it." And it still may be.'

Amodei's Pivot on Automation

Amodei's evolution has been equally dramatic. The Anthropic CEO once claimed AI could eliminate 50% of white-collar jobs, but he's now reframing automation as a productivity multiplier rather than a job destroyer. 'If you automate 90% of the job, then everyone does the 10% of the job,' he explained earlier this month. That remaining 10%, he argues, expands to fill the gap—creating something like '10-times their productivity.' The shift echoes theories from economists Alex Imas and Tyler Cowen about how automation historically creates more work rather than less.

Solomon's Quiet Vindication

Meanwhile, Goldman Sachs CEO David Solomon hasn't needed to change his position because he never held the apocalyptic one. In a New York Times op-ed, Solomon pointed to over a century of American economic history as evidence: civilian U.S. employment has grown 145% since 1962, and Goldman Sachs research shows data center construction alone added 200,000 jobs since 2022. 'Do any of us feel like we have less to do these days despite the convenience of Excel, email or Zoom?' he asked. A 2018 study by Nobel laureate Daron Acemoglu backs this view, finding AI's displacement effect is typically offset by productivity-driven demand for labor.

The Data Tells a Murky Story

The numbers offer a complicated picture. Tech layoffs through May 2026 have already passed 115,000, approaching the 124,000 logged in all of 2025—with Meta, Amazon, and Snap among those citing AI as a driver. Yet the Yale Budget Lab found no significant changes in occupational mix or unemployment duration in high-AI-exposure jobs since ChatGPT launched in late 2022. Apollo's Torsten Slok points to professions like call center employees and radiologists—roles widely predicted to be automated first—as remaining steady or growing despite wider AI adoption. 'Lower cost per interaction does not mean fewer interactions,' Slok noted. 'It means more customers served, more channels opened and more markets worth reaching.'

The Jevons Paradox Connection

Business leaders have started coalescing around an explanation: automation doesn't shrink demand for labor—it expands it. Box CEO Aaron Levie called this the 'Jevons paradox' in a LinkedIn response to Solomon's op-ed, referencing English economist William Stanley Jevons, who observed that more efficient coal burning from the Watt steam engine actually increased total coal consumption because demand surged as costs dropped. Amodei has invoked the same theory. The pattern suggests AI will deliver 'the same value proposition, but cheaper'—which historically means more work, not less.

Key Takeaways

  • Altman admitted being "pretty wrong" about AI eliminating entry-level white-collar jobs, citing his own failed experiment outsourcing Slack and email to AI
  • Amodei shifted from predicting 50% white-collar job elimination to framing automation as a productivity multiplier that expands the remaining work
  • Both OpenAI and Anthropic are reportedly preparing IPOs with $1 trillion valuations this year
  • Goldman Sachs CEO Solomon pointed to 145% civilian employment growth since 1962 as evidence against AI job apocalypse claims

The Bottom Line

The AI industry's sudden discovery that robots won't eat all our jobs after all feels less like intellectual honesty and more like pre-IPO reputation management. When your $1 trillion valuation depends on investors believing you're building the future rather than destroying it, suddenly 'we were wrong' becomes a lot easier to say.