“It is very sad for me,” Altman wrote in a message to the company’s employees about Sutzkever, before Laika’s departure was reported. “Ilya is one of the great minds of our generation, a beacon in our field, and a dear friend. His wisdom and vision are well known; his warmth and compassion are less well known, but just as important.” Dr. Jacob Pachucki, one of the company’s leading researchers, will replace Sutzkever as its chief scientist.
Sutzkever (37) is an Israeli born in the former Soviet Union who immigrated to Jerusalem at the age of 5. He began his academic studies at the Open University, but completed all his degrees at the University of Toronto, where he earned a doctorate in machine learning under the guidance of Prof. Jeffrey Hinton, one of the early pioneering researchers in the field of artificial intelligence (AI).
After a postdoctoral training at Stanford University, Sutzkever was a co-founder with Hinton of the artificial intelligence company DNNResearch. The company was acquired by Google in March 2013, only four months after it was established as a spin-off of Hinton’s research group. Sutzkever worked as a research scientist in the Google Brain team, which focused on basic research in the field of AI, until the end of 2015 when he retired to found OpenAI.
In the new company, Sutzkever led the research that led to the creation of groundbreaking models, algorithms and services such as Dall-E and ChatGPT, which are directly responsible for the GenAI revolution that is currently underway. He is considered one of the figures closest to Altman, and even accompanied him during his visit to Israel in June 2023 and spoke alongside him in his public appearances.
However, he was also one of the four board members who removed Altman from his position as CEO of the company last November, this against a background of disagreements that have not been fully clarified. These are probably related to the conflict that arose between Altman and Sutzkever regarding the speed with which OpenAI develops and distributes new models and services. Altman) were made as part of the board’s duty, which is to ensure that OpenAI builds artificial general intelligence (AGI) that benefits all of humanity,” Sutzkever told employees in an emergency meeting shortly after Altman’s ouster.
Outwardly, the two broadcast that their relationship is normal. However, according to the “New York Times”, Sutzkever did not actually return to work at the company, and his replacement, Pachuki, has since then actually held the position of chief scientist. According to sources, Altman tried to promote Sutzkever’s return, but these efforts have now officially failed with the announcement of his retirement. “Without him, OpenAI would not be what it is,” Altman wrote in X. “I am grateful for what he did here and committed to finishing the mission we started together.” Sutzkever wrote: “I am confident that OpenAI will build a safe and useful AGI. I had the honor and the privilege of working together (with Altman and the other executives, a.k.a.) and I will miss them all very much. Goodbye, and thank you for everything.”
The meaning of his departure is mainly symbolic, since in practice his involvement in the company since the impeachment attempt was nil, according to reports, and it is likely that Altman had already been preparing for his retirement for months and was preparing the company for the transition. Although Sutzkever’s qualities and abilities are exceptional and unquestioned, no one is irreplaceable, and OpenAI has no shortage of talented scientists, engineers, and developers.
His replacement Pachuki will have no difficulty stepping into his shoes, as he has already been in practice for the past six months, and even before that he was one of the company’s leading scientists and served as the research director. Pachucki has a bachelor’s degree in computer science from the University of Warsaw, a doctorate in computer science from Carnegie Mellon University (one of the world’s leading research institutions in the field of artificial intelligence), and post-doc training at Harvard University. While studying at Carnegie Mellon, Pachuki also worked as a software engineering intern at Meta.
Pachuki is considered close to Altman, and after his short-lived ouster, he was among the first four executives to announce their resignations from the company in response. Mutual loyalty is critical to the working relationship among the senior figures in the company. Her absence in the relationship between Altman and Sutzkever due to the latter’s involvement in the impeachment attempt was perhaps a major factor in his decision to retire. Pachuki’s loyalty, on the other hand, will greatly help the day-to-day operations among the top management. At the same time, it cannot be ignored that despite his many experience and abilities, Pachuki did not enjoy the same image, status and position of influence that Sutzkever had, especially in light of his status as a founder. This means that Altman further strengthens his hold on the company and his position as the most influential factor in it. In this context, Sutzkever may not be the last key figure to leave in the near future.
Unlike Sutzkever, Laika announced his departure in a much more laconic manner and announced simply with an X: “I resigned.” As of the time of writing these lines, the move has not been publicly addressed by Altman.
Last September, Leica was included in the list of the 100 most influential people in the field of AI by “Time” magazine. This, against the background of his leading, alongside Sutzkever, the Superalignment team of OpenAI. The team, established in July 2023, aims to promote scientific and technological breakthroughs that will allow humanity to control artificial intelligence. The creation of the team was a response to concerns raised in the months after the ChatGPT revelation, according to which more advanced GenAI models may have abilities similar to those of human intelligence, or more advanced than it, and their activity could cause significant damage to the human race to the point of rendering it irrelevant or destroying it.
The team, jointly led by Leica and Sutzkever, aimed to present an adequate response to these concerns within four years, when the company committed to dedicating 20% of its computing resources to this. About a month ago, Leica announced 50 grants on behalf of the team, in a total amount of nearly 10 million dollars, which were awarded to universities, research students, non-profit organizations and independent researchers who developed scientific and technological tools in order to create mechanisms that would enable control over artificial intelligence.
It is likely that his quick departure a few hours after Sutzkever is not disconnected from the retirement of the chief scientist, but one can only guess the reasons for this. It may simply be that Leica is joining Sutzkever in his new venture in a senior position, but it is also possible that following Sutzkever’s retirement, OpenAI made dramatic changes to the Superalignment team, or actually decided to close it, in a way that did not leave Leica a significant role in the company and the only choice left for him is to resign. Sutzkever was among the prominent voices in OpenAI for the responsible development of artificial intelligence, and his concerns in this regard were also part of the motivations that led to his support for the move to oust Altman. Now, with his retirement, Altman may have decided to reduce an activity he perceives as unnecessary and the Superalignment team was one of the main victims of this move.
ظهرت في الأصل على www.calcalist.co.il