After a string of blockbuster model releases, fundraising at a reported valuation approaching $1 trillion, and an annual run rate that’s nearly parabolic, Anthropic
has now hired one of its rival OpenAI’s most famous alumni.
“Personal update: I’ve joined Anthropic,” Andrej Karpathy
wrote on X on Tuesday, adding that the next few years at the frontier of LLMs would be “especially formative” and that he was eager to get back to research. He started this week.
It’s a major move for Karpathy, whose writing on AI is followed by nearly 2 million people on X. He was first a founding member of OpenAI in 2015, left to run AI at Tesla, came back in 2023, and left only a year later to start his own education company, Eureka Labs.
The thing—or phrase—that etched Karpathy into AI legend, though, was a viral
social media post from last year describing a “new kind of coding” that he called “
vibe coding.”
As an Anthropic employee, Karpathy will be building on the work from another popular post about a concept he called “autoresearch,” in which he found that tweaks applied to a small AI model running unsupervised could be applied to a larger model to successfully cut training time.
Teaching that method looks, more or less, like Karpathy’s new job. According to Anthropic, he will be starting a team focused on using Claude to accelerate pretraining research, the large-scale training runs that give Claude its core knowledge and capabilities.
—Eva Roytburg
