Sébastien Bubeck
Researcher in machine learning · Academics · Born 1985 · age 41
The mathematician who spent years refining the theory of how machines learn from choices, then turned that precision on the black box itself - cracking open transformer models to see what makes them tick.
Sébastien Bubeck is read about 137 times a day.
A typical day, 30 days to 1 November 2026
The middle day of the days counted, so one week of news cannot move it. The line under it is the last 400 days, one point a day.
Something wrong?Media & music
Photos 1
On YouTube
24.5K subscribersAbout this person
Born April 16, 1985, Sébastien Bubeck built his early career at the edges of theory and practice: Berkeley, then Princeton as assistant professor, working the problems of online learning and optimization. Microsoft Research Redmond brought him in to lead the Machine Learning Foundations group, and he rose to vice president of applied research and distinguished scientist. The work shifted as the models grew larger - he trained the theoretical lens on deep neural networks, particularly transformers, asking questions about why they work at all. In 2024 he moved to OpenAI, carrying that same scrutiny forward.
