One of the most prominent minds in artificial intelligence is switching teams once again. Andrej Karpathy, the highly respected AI researcher who co-founded OpenAI and spent years leading the computer vision team at Tesla, has officially joined Anthropic.
The announcement came directly from the researcher on social media on Tuesday. “Personal update: 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,” Karpathy shared on X. He added that while he is jumping back into core research, he remains deeply passionate about education and plans to resume his work on it in time.
Accelerating Pre-Training Research
Karpathy is joining Anthropic’s pre-training team, working under team lead Nick Joseph. Pre-training is the incredibly expensive, compute-heavy phase where a large language model like Claude builds its foundational knowledge before fine-tuning.
According to Anthropic, Karpathy will be building a brand-new team inside the company that focuses entirely on using Claude itself to accelerate pre-training research. Tapping a researcher of his caliber to lead an AI-assisted research team signals that Anthropic wants to use clever architectural optimization, rather than just throwing infinite computational bandwidth at the problem, to stay competitive with rivals like OpenAI and Google.
A Look Back at Karpathy’s Legacy
Before his recent transitions through the LLM space, Karpathy was a cornerstone of Tesla’s automation journey. He served as Tesla’s Director of AI, leading the Autopilot and Full Self-Driving computer vision teams during a critical era of development. Karpathy announced his departure from Tesla in the summer of 2022 to take a break and focus on personal projects.
Since leaving the automaker, Karpathy has frequently expressed interest in working at Tesla again, specifically mentioning that he would love to contribute to the company’s Optimus humanoid robot project. Even Elon Musk publicly invited Karpathy to return to the company last summer, noting that he would love to work with his former AI chief again.
Instead, Karpathy returned to OpenAI for a year before departing in early 2024 to launch Eureka Labs, an AI-powered education startup. While his new role at Anthropic puts Eureka Labs on the back burner, it highlights just how rapidly the landscape of artificial general intelligence is shifting. Karpathy has often offered fascinating glimpses into the future of AGI, viewing it as an inevitable transition in human technology.
Anthropic Continues Its Talent Hunt
Karpathy isn’t the only major hire Anthropic secured this week. The firm also brought on cybersecurity veteran Chris Rohlf to join its frontier red team. Rohlf, who spent six years at Meta and previously worked on the well-known “The Paranoids” security team at Yahoo, will be tasked with stress-testing Anthropic’s advanced models against severe digital threats.
By securing both top-tier machine learning pioneers and robust security experts, Anthropic is solidifying its position at the absolute frontier of AI development. It remains to be seen how Karpathy’s unique background in physical computer vision at Tesla will shape the future of text and multimodal models like Claude, but his presence alone makes Anthropic a company to watch closely over the next few months.