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

Andrej Karpathy, a prominent figure in the AI community known for his contributions as a co-founder of OpenAI and his leadership role at Tesla, has officially joined Anthropic. Announcing his new position on X, Karpathy expressed his enthusiasm, stating, "I’ve joined Anthropic. I think the next few years at the frontier of LLMs will be especially formative. I am very excited to join the team here and get back to R&D." Starting this week, Karpathy will focus on pre-training efforts under the guidance of team leader Nick Joseph. Pre-training is a critical phase in AI development, responsible for large-scale training that forms the backbone of models like Claude. This process is not only resource-intensive but also crucial for enhancing AI capabilities. According to an Anthropic spokesperson, Karpathy's role will involve establishing a dedicated team aimed at utilizing Claude to expedite pre-training research. Karpathy's unique ability to connect theoretical aspects of large language models (LLMs) with practical training applications positions him as a key asset for Anthropic. His recruitment signals the company's commitment to advancing AI-assisted research, a strategy intended to keep pace with competitors such as OpenAI and Google. Prior to his time at Anthropic, Karpathy was instrumental at OpenAI, focusing on deep learning and computer vision until 2017, when he transitioned to Tesla. At Tesla, he oversaw the Full Self-Driving and Autopilot programs until his departure in 2022. After a brief return to OpenAI, he founded Eureka Labs, a startup aimed at integrating AI into educational tools. While he has maintained a presence in education through online courses and lectures, the future of Eureka Labs remains uncertain as he embarks on this new chapter. In addition to Karpathy's onboarding, Anthropic has welcomed Chris Rohlf to its frontier red team, which is tasked with stress-testing AI models against potential cyber threats. Rohlf, a cybersecurity veteran with over two decades of experience, previously worked with Yahoo's cybersecurity team and held a position at Meta. He expressed optimism about the potential for AI to significantly enhance cybersecurity measures, stating, "We have a real opportunity in front of us to dramatically improve cybersecurity with AI. I can’t think of a better company or team to join at this critical moment in time."