Sam Altman and Dario Amodei Reverse Job Doom Narrative

For two years, the loudest voices in Silicon Valley warned that Gen AI could dismantle white-collar work. Now the message has shifted.

Sam Altman, CEO at OpenAI, and Dario Amodei, CEO at Anthropic, are stepping back from doomsday forecasts. Both now frame automation as a productivity boost rather than a headcount reducer.

The pivot arrives as both companies prepare for public listings in 2026 as investor optics and regulatory scrutiny are at front of their mind. The underlying labour market, on the other hand, remains steadier than the hype cycle suggests.

From alarm to admission

For much of the early Gen AI cycle, prominent leaders predicted widespread white-collar displacement where entry-level roles appeared especially exposed.

Speaking with Matt Comyn, CEO at Commonwealth Bank of Australia, in May 2026, Sam concedes that the expected short-term shock has not appeared. He says: “I am delighted to be wrong about this. I thought there would have been more impact on entry-level white-collar jobs being eliminated by now than has actually happened.”

Dario also reframes the impact, describing automation as a multiplier of output rather than a destroyer of roles. He says: “If you automate 90% of the job, then everyone does the 10% of the job. And the 10% kind of expands to be 100% of what people do and kind of 10-times their productivity.”

This tone shift brings the tech narrative closer to the language of enterprise efficiency. It replaces existential risk with incremental gains.

In doing so, it aligns with a century of economic history rather than a break from it.

Valuations and regulatory optics

OpenAI and Anthropic are reportedly targeting US$1tn valuations in planned IPOs in 2026. With such listings at stake, stability is the sale.

Institutional investors favour predictable growth, enterprise resilience and regulatory compliance. Few want exposure to a technology which could trigger a labour crisis.

There is also a regulatory dimension. Framing Gen AI as an employment threat invites stronger antitrust and labour market scrutiny.

By recasting the technology as a corporate utility and not an economy-wide disruptor, they are reducing both headline and policy risk.

This approach mirrors the posture of financial leaders like David Solomon, CEO at Goldman Sachs. He has maintained since late 2025 that the panic is overstated.

Wall Street’s language, not science fiction

David draws a parallel from electrification in the 1900s to the digital revolution of the 1990s. He says: “The United States has a long track record of creating new jobs in response to disruption.” “I do not see any reason to think this dynamic will stop now.”

The message to markets is clear. Gen AI is a lever for productivity, not a fuse for mass unemployment.

Investors, particularly pension funds and asset managers, prefer that framing. It offers earnings uplift without social upheaval.

Executives attuned to IPO roadshows speak in that register to reduce volatility in both perception and valuation.

Jobs data and Jevons paradox

Tech layoffs exceed 115,000 through May 2026, yet occupational unemployment remains broadly unchanged, according to the Yale Budget Lab. The disruption is sector specific, not systemic.

Economists point to Jevons paradox which postulates that when a resource becomes more efficient and cheaper, demand for it tends to rise.

Software follows the same pattern. Automating a task lowers its cost, which expands demand for the output of that task.

Aaron Levie, CEO at Box, highlights this logic. He asks why the acceleration of work in recent decades has not eliminated jobs. 

He says: “If you looked at what work looked like a few decades ago and saw how much faster everything is or easier it is to produce today, even before AI, you’d certainly have been convinced there’d be no jobs left. Yet the opposite has happened. Why?”

The labour market absorbs technology by reshaping roles rather than erasing them outright.

Aggregate resilience can mask individual strain. The 90/10 split that Dario describes looks efficient in a chart. However, in practice, routine tasks provide cognitive buffer time which supports sustained focus.

When automation strips those away, work compresses into a sequence of complex decisions. Workers often do not receive time back. The baseline of expected output rises instead, as David notes.

As the frontier shifts from headcount to human attention, managing cognitive load becomes a core design question for tools and teams.

The Gen AI narrative has moved from apocalypse to efficiency, timed with landmark IPOs. Markets and regulators welcome the shift.

Nevertheless, one can say that white-collar employment proves more resilient than early forecasts suggest. However, productivity gains carry a human cost if not managed well.