OpenAI loses its biggest names in AI safety as Ilya Sutskever and Jan Leike walk away
- Jan Leike, researcher at Super AI Alignment, also departs
Update:
A few hours after Sutskever’s resignation, AI safety researcher Jan Leike also announced his departure. Leike’s main focus is the alignment of superintelligent AI for the benefit of humanity, so-called alignment research.
Along with Sutskever, Leike was co-lead of the OpenAI Superalignment team, founded in summer 2023 with the goal of iteratively aligning superintelligence and creating an automated alignment researcher with human-like capabilities.
Leike does not give a reason for his resignation. Around the time of Altman’s firing (see below), there were rumors of fundamental disagreements within the company over commercial focus versus research focus and the associated AI risks that come with commercialization.
Ad
Ad
Original article from May 15, 2024:
Ilya Sutskever, co-founder of OpenAI and one of the authors of the seminal 2012 AlexNet paper that marked the breakthrough of deep learning, is leaving the company after nearly a decade.
Sutskever described OpenAI’s trajectory as “miraculous” and said he is confident the company will achieve its ultimate goal of creating safe and useful artificial general intelligence (AGI).
Sutskever plans to focus on a new project that is “personally meaningful” to him, but didn’t provide details.
His successor as head of research will be Jakub Pachocki, a seven-year veteran of OpenAI who considers Sutskever his mentor.
Recommendation
OpenAI CEO Sam Altman calls Pachocki “easily one of the greatest minds of our generation” and says he is confident in his ability to lead the company quickly and safely towards AGI. According to Altman, Pachocki has already led “many” of OpenAI’s “most important projects.”
Altman also had warm words for Sutskever, saying, “I am forever grateful for what he did here and committed to finishing the mission we started together.”
Sutskever’s OpenAI rollercoaster
Sutskever’s departure follows a tumultuous time at OpenAI. As a board member, he was involved in the brief ouster of CEO Sam Altman in November, reportedly over concerns about the rapid commercialization Altman was pushing and the security risks it posed.
An investigation by the law firm WilmerHale concluded that Altman’s actions did not warrant dismissal and that the board’s decision was premature.
Sutskever apologized for his role in the incident and helped reinstate Altman before resigning from the board and disappearing from public appearances.
