Anthronyone with a microphone that AI is about to wipe out half of all entry-level white-collar jobs. Now, at the World Economic Forum in Davos, the man whose company is selling that very AI is no longer just warning about the layoffs. He’s calling the policy response to them a necessity that will eventually steamroll partisan politics in the United States.In an interview with WSJ Editor-in-Chief Emma Tucker, Amodei argued the country won’t have the luxury of arguing about whether to step in. “We’re going to find that ideology will not survive the nature of this technology. It won’t survive reality,” he told the WSJ. The interventions he’s been pushing—government-backed economic mobility programs, redistribution, retraining schemes—”can become bipartisan and universal because everyone will recognize the necessity of it. Just mark my words.”It’s a striking position from a CEO who is already being accused of sounding like a doomsayer. Amodei has told Axios that unemployment could hit 10-20% within five years. He’s said coding jobs go first, then finance, then law and consulting. At Anthropic’s financial services briefing in New York on May 5, sitting next to JPMorgan CEO Jamie Dimon, he said software itself could become “essentially free” and that “whole jobs, whole careers” might not survive.
Anthropic’s $900 billion valuation runs on the doomsday pitch
The framing is hard to separate from the company’s IPO arithmetic. Anthrouation north of $900 billion, with annual revenue running above $40 billion. To justify that math, institutional investors—pension funds, sovereign wealth, the BlackRocks of the world—need to believe Claude isn’t a productivity add-on but a replacement for human labour at industrial scale. The bigger the displacement story, the bigger the addressable market.Anthropic’s own researchers quietly complicate the script. A March 2026 paper from the company found Claude is currently covering just 33% of tasks in the computer and math category, against a theoretical capability of 94%. The same data showed no systematic rise in unemployment among the most AI-exposed workers. The only real warning sign was a 14% drop in hiring of 22-to-25-year-olds into exposed roles since ChatGPT launched. Concerning. Not quite the end of white-collar work.
Mississippi, not Silicon Valley: the geography Amodei wants Washington to fix
Where Amodei’s argument lands more interestingly is on geography. In the WSJ sit-down, he flagged the gap between AI’s economic centre of gravity and the rest of the country. “How do we get economic growth in Mississippi? That is coming this contained area of Silicon Valley,” he said, before tying that gap to work Anthroshington should write checks because AI is evil. It’s that the geography of AI gains will be so lopsided that resistance to intervention won’t hold.Plenty of his peers aren’t buying it. Yann LeCun, Meta’s former chief AI scientist, has said Amodei “knows absolutely nothing about the effects of technological revolutions on the labour market.” Google DeepMind CEO Demis Hassabis told WIRED this week that companies cutting engineers because of AI suffer from “a lack of imagination.” Sam Altman, who spent early 2026 matching Amodei’s register, has quietly pivoted to talking about augmenting workers rather than replacing them.Amodei isn’t budging. Whether the bipartisan consensus he’s predicting actually shows up—or stays a useful talking point on the IPO roadshow—is something Washington and Wall Street will have to figure out before his five-year clock runs out.
