Last July, OpenAI revealed the establishment of a new research team that would be getting ready for the emergence of highly intelligent artificial intelligence that could potentially surpass its creators. Ilya Sutskever, OpenAI’s chief scientist and co-founder, was appointed as the co-lead of this new team. OpenAI stated that the team would have access to 20% of its computing power.
However, OpenAI has now confirmed that the “superalignment team” no longer exists. This decision follows the departure of several researchers, the news of Sutskever leaving the company, and the resignation of the team’s other co-lead. The group’s work will now be integrated into OpenAI’s other research initiatives.
Sutskever’s departure made headlines as he played a key role in helping CEO Sam Altman launch OpenAI in 2015 and shaping the research that led to ChatGPT. Despite this, Sutskever was one of the four board members who dismissed Altman in November. Altman was reinstated as CEO after a tumultuous five days, following a revolt by OpenAI employees and a deal that saw Sutskever and two other board members stepping down.
Shortly after Sutskever’s departure was announced, Jan Leike, the former DeepMind researcher who co-led the superalignment team, announced his resignation.
Both Sutskever and Leike did not respond to requests for comment. Sutskever did not provide a reason for his departure but expressed confidence in OpenAI’s current direction in a post. He stated, “The company’s progress has been remarkable, and I believe that OpenAI will develop AGI that is safe and beneficial.”
Leike explained in a post that his decision to leave stemmed from disagreements over the company’s priorities and the allocation of resources to his team.
The disbandment of OpenAI’s superalignment team is another sign of internal changes within the company following last November’s governance crisis. Two team members, Leopold Aschenbrenner and Pavel Izmailov, were reportedly dismissed for leaking company information. William Saunders, another team member, left OpenAI in February, according to an online forum post.
Additionally, two other OpenAI researchers focusing on AI policy and governance seem to have departed the company recently. Cullen O’Keefe resigned from his role in April, and Daniel Kokotajlo left due to concerns about the company’s approach to AGI, according to online forums. These researchers did not respond to requests for comment.
