OpenAI co-founder Andrej Karpathy joins Anthropic's pre-training team
TechCrunch • • general
Key Points:
- Andrej Karpathy, former AI lead at Tesla and OpenAI co-founder, has joined Anthropic to work on pre-training large language models (LLMs), focusing on accelerating research using Anthropic's Claude model.
- Karpathy will lead a new team at Anthropic aimed at leveraging AI-assisted research to improve pre-training efficiency, highlighting the company's strategy to compete with OpenAI and Google beyond just compute power.
- Previously, Karpathy led Tesla’s Full Self-Driving and Autopilot programs, returned briefly to OpenAI, and founded Eureka Labs, an AI education startup, though his current involvement with Eureka Labs is unclear.
- Anthropic also hired cybersecurity veteran Chris Rohlf to its frontier red team, which tests AI models against security threats; Rohlf brings over 20 years of experience from Yahoo, Meta, and Georgetown’s Center for Security and Emerging Technology.
- Karpathy expressed ongoing passion for AI education and plans to resume related work in the future, complementing his research and development role at Anthropic.