Jan Leike, the co-head of OpenAI’s Superalignment team, recently resigned from his position, citing disagreements with the company’s leadership over its core priorities. Leike, who had been with OpenAI since its inception, played a significant role in the development of several AI models, including ChatGPT, GPT-4, and InstructGPT.
In a series of posts on X (formerly Twitter), Leike expressed his concerns about the company’s focus. He believes that OpenAI should prioritize security, monitoring, preparedness, safety, adversarial robustness, alignment, confidentiality, societal impact, and related topics. He also stated that he had reached a “breaking point” due to these disagreements.
Leike’s departure comes at a time when OpenAI is grappling with internal turmoil. The Superalignment team, which was responsible for managing and guiding superintelligent AI systems, has reportedly been dissolved. The team had been promised 20% of the company’s compute resources, but their requests for even a fraction of that were often denied, leading to internal disputes and hindering their work.
Leike’s resignation and the dissolution of the Superalignment team raise questions about OpenAI’s commitment to the safety and control of superintelligent AI systems.
Read more: mashable.com