Why the AI Doomsayers Suddenly Love Developers Again

A year ago, Sam Altman and Dario Amodei were the loudest voices in the room predicting that AI would erase half of white-collar work. In May 2026, both are suddenly softening the message. The pivot is too clean to be a coincidence — and the timing tells you everything.

From “We’ll Replace Devs” to “We Love Devs”

A Hacker News thread on May 28 (63 points, 51 comments) cut to the heart of it. The title was blunt: both Altman and Amodei are walking back their AI jobs apocalypse predictions.

The top comment nailed the absurdity in a single line: going from “we will replace developers” to “we love developers, please keep burning tokens” is, in fact, pretty funny.

Last year, Amodei told Axios that AI could eliminate half of all entry-level white-collar jobs within five years. Altman has spent two years repeating that agents would soon “replace employees.” In recent interviews, the talking points have shifted to something much gentler: AI is a tool, AI augments human work, AI raises the ceiling rather than removing the floor.

Why the Pivot, Why Now

The timing isn’t subtle. OpenAI is widely expected to start lining up an IPO in the second half of 2026. Anthropic is reportedly preparing another mega-round. Apocalypse marketing plays well in a Series whatever pitch deck. It plays terribly in an S-1.

Think like an institutional investor for a second. How does the SEC react to a CEO who has spent two years telling reporters his product will gut the labor market? How does the EU, mid-rollout of the AI Act and actively scoping labor-impact assessments, treat a company whose own founder has handed regulators the quote sheet? “Existential job destroyer” is the kind of self-description compliance teams have nightmares about.

There’s a B2B angle too. Enterprise buyers — the ones actually signing seven-figure contracts — find it much easier to greenlight “a tool that makes our employees more productive” than “a tool that makes our employees redundant.” Less union friction, less internal political cost, cleaner procurement story. The augmentation narrative just sells better once you’re chasing Fortune 500 logos instead of VC term sheets.

The Community Isn’t Buying It

The HN reaction is icy. One commenter put it cleanly: confidently making sweeping claims for years and then frantically retreating the moment public sentiment turns just reinforces the impression that these people are not to be trusted. Another comment was even shorter: the damage is done.

And that’s the real point. Hiring freezes were justified internally with slide decks quoting these exact CEOs. New-grad pipelines collapsed. Bootcamps imploded. Even CS enrollment trends have been read through the lens of “why train for a job AI will eat.” None of those decisions get reversed because Altman now says, on reflection, that humans are pretty great.

The Real Cost of the Rhetoric

What this episode actually exposes is something more uncomfortable than a flip-flop: the dominant narrative around AI tracks the funding cycle, not the technology. Seed and growth stages need revolution. Pre-IPO stages need stability. Public markets need “responsible enabler of human productivity.” The story rotates because the audience rotates.

Watch what happens after the IPO, if it happens. Once quarterly earnings pressure kicks in and the company needs to justify a trillion-dollar valuation, the agents-replacing-everyone story is conveniently the most powerful one available. Don’t be surprised when it comes back.

So which is it — a genuine reassessment, or roadshow theater with a six-month half-life? And if it’s the second one, the more interesting question is who has been paying the bill for the last two years of apocalypse talk. It hasn’t been the people writing the press releases.