Andrej Karpathy has joined Anthropic to return to frontier LLM R&D, while stating plans to resume education work in time.
Key Takeaways
Karpathy left independent work to join Anthropic, citing the next few years at the LLM frontier as “especially formative.”
His stated focus is R&D, not a named role or team.
He explicitly intends to continue education work eventually, suggesting it is paused, not abandoned.
EurekaLabs, his education startup, appears effectively on hold; its X account posts are now restricted.
Hacker News Comment Review
Commenters debated whether Karpathy will do meaningful frontier research or function primarily as a high-profile communicator and marketer for Anthropic and Claude.
Concern surfaced that his public educational output, widely considered best-in-class for LLM fundamentals, may now disappear into a private internal void.
His career arc (OpenAI, Tesla, independent, Anthropic) reads to many as someone who has consistently just missed major inflection points at each stop, raising questions about timing and fit.
Notable Comments
@ryeguy_24: He foreshadowed this in a recent interview, saying frontier labs could re-engage him if he risked falling out of touch with evolving approaches.
@markerbrod: EurekaLabs X account posts are now restricted, signaling the venture is likely wound down.
