Andrej Karpathy, the AI researcher who co-founded and previously labored at OpenAI and beforehand led AI at Tesla, has joined Anthropic.
“I’ve joined Anthropic,” Karpathy posted on X Tuesday. “I believe the subsequent few years on the frontier of LLMs shall be particularly formative. I’m very excited to affix the workforce right here and get again to R&D.”
Karpathy began this week at Anthropic, the place he’s engaged on pre-training beneath workforce lead Nick Joseph. Pre-training is accountable for the large-scale coaching runs that give Claude its core information and capabilities, in keeping with the corporate. It’s additionally one of the crucial costly, compute-intensive phases of constructing a frontier mannequin.
An Anthropic spokesperson instructed TechCrunch that Karpathy will begin a workforce centered on utilizing Claude to speed up pre-training analysis.
Karpathy is likely one of the few researchers who can bridge the hole between LLM concept and large-scale coaching follow. Tapping him to construct such a workforce is a transparent signal from Anthropic that it believes AI-assisted analysis, moderately than pure compute, is the way it stays aggressive with OpenAI and Google.
Whereas at OpenAI, Karpathy centered on deep studying and laptop imaginative and prescient till he departed in 2017 to affix Tesla. He led Tesla’s Full Self-Driving (FSD) and Autopilot applications earlier than leaving in 2022.
He then went again to OpenAI for one 12 months earlier than leaving once more in 2024 to start out Eureka Labs, a startup devoted to making use of AI assistants to training.
Karpathy hasn’t shared many updates on Eureka Labs since its launch, and it’s not clear if the famend researcher will proceed with the startup. He has additionally taught an internet course referred to as Neural Networks: Zero to Hero that helps college students study to construct neural networks from scratch in code, and has a YouTube channel the place he semi-regularly posts lectures on LLMs and AI.
“I stay deeply enthusiastic about training and plan to renew my work on it in time,” Karpathy stated.
TechCrunch has reached out to Karpathy for remark.
Individually, Anthropic has additionally brought on Chris Rohlf to its frontier crimson workforce, which stress-tests superior AI fashions towards extreme threats. Rohlf is a veteran of the cybersecurity trade with greater than 20 years of expertise. He beforehand labored at Yahoo’s well-respected cybersecurity workforce often called “The Paranoids,” and extra just lately at Meta, the place he labored for six years earlier than becoming a member of Anthropic. Rohlf was additionally a fellow at Georgetown’s Middle for Safety and Rising Know-how, the place he labored on the CyberAI undertaking.
“Now we have an actual alternative in entrance of us to dramatically enhance cyber safety with AI,” Rohlf stated in a post on X. “I can’t consider a greater firm or workforce to affix at this vital second in time.”
Once you buy by way of hyperlinks in our articles, we may earn a small commission. This doesn’t affect our editorial independence.
