Andrej Karpathy, one of OpenAI’s founding members and the former director of AI at Tesla Inc (NASDAQ:TSLA), announced Tuesday morning that he is joining Anthropic.

“I think the next few years at the frontier of LLMs will be especially formative,” Karpathy wrote, adding that he plans to return to hands-on research.

Karpathy left OpenAI in February 2024 to launch education startup Eureka Labs, but the destination tells the real story. Sam Altman’s lab keeps losing senior alumni to Dario Amodei’s.

A Pattern, Not An Anecdote

Karpathy is the third high-profile OpenAI figure to land at Anthropic in under two years. Jan Leike, OpenAI’s former head of alignment, defected in May 2024. Co-founder John Schulman followed that August.

The flow has only gone one way. No comparable Anthropic researcher has moved to OpenAI in that window.

Karpathy is also one of the most-followed voices in AI, with his “vibe coding” essays and Claude Code workflow posts routinely driving developer discourse.

A Karpathy-inspired CLAUDE.md repository has racked up over 220,000 GitHub stars since January.

Polymarket Already Priced The Race

Traders on Polymarket assign Anthropic a 70% probability of having the best AI model at the end of June, based on the Chatbot Arena leaderboard. OpenAI 5%. That contract has logged $6 million in volume.

The IPO race is even more lopsided. Polymarket gives Anthropic a 67.5% chance of going public before OpenAI, with the Claude maker reportedly in talks for a $30 billion round at a $900 billion valuation. That would surpass OpenAI’s $852 billion mark from its March raise.

OpenAI has other problems, despite winning its $180 billion lawsuit against Elon Musk, it has missed user growth targets earlier this year, and its CFO has openly worried about paying for compute.

Karpathy’s move adds a marquee name to a roster that is already winning the talent war and, according to traders, the model race.

Image: Shutterstock