Andrej Karpathy (안드레이 카파시), an AI expert who co-founded OpenAI and previously served as Tesla's head of AI, has joined Anthropic, TechCrunch reported on May 19 local time.

Karpathy said on social media platform X (Twitter), "I joined Anthropic." He added, "The LLM field will be an especially important period over the next few years. I am very excited to join the Anthropic team and return to R&D work."

Karpathy has been working since this week at Anthropic on the pre-training team led by Nick Joseph. Pre-training is one of the most expensive stages in building frontier models.

An Anthropic spokesperson said Karpathy will create a new team that uses Claude to accelerate pre-training research.

Karpathy is considered one of the few researchers who can cover both LLM theory and hands-on work in large-scale training.

TechCrunch said Anthropic's hiring of Karpathy signals it will compete with OpenAI and Google by using AI to speed up research rather than pouring in more computing resources. Karpathy researched deep learning and computer vision at OpenAI before moving to Tesla in 2017. He led Tesla's Full Self-Driving (FSD) and Autopilot programs and left the company in 2022. He later returned to OpenAI for a year before leaving again in 2024 to found AI education startup Eureka Labs.

It is unclear whether Karpathy will continue his work at Eureka Labs. Karpathy said, "My passion for education remains unchanged, and I plan to resume this work when the time comes." Karpathy is also known as the creator of the concept of vibe coding, which uses AI to code in natural language.

Anthropic also hired cybersecurity industry veteran Chris Rolfe for its frontier red team. Rolfe has more than 20 years of experience and worked on Yahoo's security team 'The Paranoids' and spent six years at Meta.