Axios reports, citing people familiar with the matter, that Anthropic CEO Dario Amodei is worried some of the top talent joining the company are drawn by big salaries and equity rather than Anthropic's AI safety mission.
His concern isn't exactly baseless. Top AI researchers are hopping between the big labs. Thinking Machines has had four co-founders leave within a year. Google's Noam Shazeer and John Jumper have each left — Shazeer to OpenAI, Jumper to Anthropic. And some researchers Meta spent heavily to poach quickly jumped ship again.
Money remains the most direct lure. Some researchers want to get equity before a potential Anthropic or OpenAI IPO. Others worry that as AI takes over more model development work, the scarcity and bargaining power enjoyed by top researchers today could erode fast.