OpenAI co-founder Andrej Karpathy joins Anthropic’s training team

Andrej Karpathy, an AI researcher who founded and worked at OpenAI and led AI at Tesla, has joined Anthropic.

“I’ve joined Anthropic,” Karpathy written on X Tuesday. “I think the next few years at the LLM frontier are going to be very developmental. I’m very excited to be joining the team here and getting back to R&D.”

Karpathy started this week at Anthropic, where he is working on a pre-event training program led by Nick Joseph’s team. The pre-training is the driving force behind the major training that gives Claude his great knowledge and potential, according to the company. It is also one of the most expensive, computer-intensive parts of building a frontier model.

An Anthropic spokesperson told TechCrunch that Karpathy will start a team focused on using Claude to improve the research it’s already trained on.

Karpathy is one of the few researchers who can bridge the gap between LLM theory and major teaching practices. Hiring him to create such a team is a clear signal from Anthropic that they believe that AI-powered research, rather than pure computing, is how they compete with OpenAI and Google.

While at OpenAI, Karpathy focused on deep learning and computer vision until he left in 2017 to join Tesla. He led Tesla’s Full Self-Driving (FSD) and Autopilot programs before leaving in 2022.
He then returned to OpenAI for a year before leaving again in 2024 start Eureka Labsintroduction to the use of AI assistants in education.

Karpathy hasn’t shared many updates on Eureka Labs since its inception, and it’s unclear if the noted researcher will continue with the startup. He has also taught online courses called Neural Networks: Zero to Hero which helps students learn how to build neural networks from scratch, and has a YouTube video where he regularly places courses on LLMs and AI.

“I am passionate about education and plan to resume my career in due course,” said Karpathy.

TechCrunch has reached out to Karpathy for comment.

Individually, Anthropic too to be brought Chris Rohlf to its red team, which tests the stress of advanced AIs against larger threats. Rohlf is a cybersecurity veteran of over 20 years. He previously worked at Yahoo’s respected cybersecurity team called “The Paranoids,” and most recently at Meta, where he worked for six years before joining Anthropic. Rohlf was also a fellow at Georgetown’s Center for Security and Emerging Technology, where he worked on CyberAI.

“We have a real opportunity in front of us to advance cyber security with AI,” Rohlf said. post on X. “I can’t think of a better company or team to join at this critical time.”