Andrej Karpathy Departs OpenAI Once More — He Assures There's No Drama Involved

Andrej Karpathy, a highly esteemed research scientist, announced today that he has departed from OpenAI. This marks the second occasion that Karpathy has moved on from the prominent AI organization, and he clarified that his departure is not due to any specific event, issue, or drama.

Karpathy, a founding member of OpenAI, first left the organization in 2017 to join Tesla, where he led the autopilot team. After leaving Tesla in 2022, he rejoined OpenAI nearly a year ago. In addition to his work in AI, Karpathy has garnered a substantial following on social media and YouTube, where he shares insightful writings and videos that delve into the evolving field of artificial intelligence.

“Hi everyone, yes, I left OpenAI yesterday. First off, nothing ‘happened,’ and it’s not a result of any specific event, issue, or drama (but please keep the conspiracy theories coming, as they are highly entertaining :)),” Karpathy tweeted. “Actually, my experience at OpenAI over the past year has been fantastic—the team is incredibly strong, the people are wonderful, and the roadmap is thrilling. I think we all have exciting times ahead. My immediate plan is to focus on my personal projects and see where that leads. Those who have followed me for a while might have an idea of what that could entail.”

Karpathy’s exit, first reported by The Information, occurs after a tumultuous year at OpenAI, marked by Sam Altman’s brief ousting as CEO before his reinstatement amidst high-profile drama. A spokesperson for OpenAI confirmed that Karpathy’s responsibilities have been transitioned to another researcher, whose name has not been disclosed.

“Andrej is leaving to pursue personal projects. We are immensely grateful for his contributions and wish him the best in his future endeavors. His responsibilities have been taken over by a senior researcher who closely collaborated with Andrej,” the spokesperson stated.

Karpathy's now-updated bio on social media previously stated he was “building kind of a J.A.R.V.I.S at @OpenAI,” suggesting he was developing an AI assistant project within the company. Earlier this year, he elaborated on his vision behind this concept. “I reference J.A.R.V.I.S in general terms because it represents one of my favorite portrayals of AI—a helpful, conversational, and empowering automation. It serves as an ally against adversity and chaos,” he explained.