Ilya Sutskever, one of the co-founders and longtime chief scientist of OpenAI, has announced his departure from the company. OpenAI CEO Sam Altman expressed his sadness at Sutskever’s departure, describing him as one of the greatest minds of their generation and a guiding light in the field of AI. Altman also credited Sutskever for his significant contributions to OpenAI and emphasized his commitment to continue the mission they started together.
Replacing Sutskever as the new chief scientist is Jakub Pachocki, who joined OpenAI in 2017 as a research lead on the Dota team. Pachocki played a crucial role in developing an AI system capable of defeating human players in Valve’s Dota 2 strategy game. He then progressed to become a research lead in OpenAI’s reasoning and science of deep learning organizations before being promoted to principal of research.
There is uncertainty about whether Pachocki will also take over as the head of OpenAI’s Superalignment team, which was previously under the purview of Sutskever and Jan Leike. Leike has also resigned from OpenAI, according to The New York Times. The Superalignment team was established to explore ways to regulate and govern “superintelligent” AI systems, which possess intelligence surpassing that of humans. John Schulman, another co-founder of OpenAI, is expected to assume the role of overseer for the team.
OpenAI’s president, Greg Brockman, expressed his gratitude to Sutskever for his instrumental role in building the foundations of the company. Brockman referred to Sutskever as his co-founder, friend, and even the officiant at his civil ceremony. Their partnership helped shape what OpenAI has become today.
Sutskever’s departure marks the end of a saga that began last November. At that time, Sutskever and OpenAI CTO Mira Murati raised concerns about Altman’s behavior and disagreements over the company’s direction. The previous board of directors, which included Sutskever, abruptly removed Altman without proper notification, causing outrage among Microsoft, OpenAI’s investors, and employees. Eventually, Altman was reinstated, leading to the resignation of most of the old board. Sutskever never returned to work after that, and Pachocki effectively took on the role of chief scientist.
Sutskever, who obtained his Ph.D. in computer science from the University of Toronto under the guidance of AI expert Geoffrey Hinton, joined OpenAI in 2015 after leaving Google Brian. His contributions to the field of AI are significant, including his involvement in developing ImageNet, one of the first modern computer vision systems, and DeepMind’s AlphaGo, an AI system capable of playing the game Go.
As for Sutskever’s future plans, he has not disclosed them yet. However, he expressed his belief that OpenAI will successfully create artificial general intelligence that is both safe and beneficial. This departure marks the end of an era for OpenAI but also signals new opportunities for growth and innovation in the field of AI.