Sebastien Bubeck, a key figure in Microsoft's AI research, is leaving the tech giant to join OpenAI. Bubeck, who served as VP of AI and Distinguished Scientist at Microsoft, was instrumental in developing the company's Phi family of small language models.
Bubeck's work at Microsoft has been pivotal in the company’s generative AI efforts, particularly in leveraging models that aim to be more efficient alternatives to OpenAI's GPT models. His research led to the development of smaller, cost-effective Phi models, some of which have even been integrated into Microsoft's Bing chatbot and Office 365 Copilot, replacing OpenAI’s models in certain areas.
It's currently unclear what specific role Bubeck will take on at OpenAI, but a Microsoft spokesperson said in a statement that he had "decided to leave Microsoft to further his work toward developing" artificial general intelligence. This transition comes as OpenAI itself has faced a wave of changes in its research ranks, including the departure of its former CTO, Mira Murati, whose last day was last Friday.
Microsoft appears unfazed by the move, expressing its intention to maintain a productive relationship with Bubeck through OpenAI, a company it has heavily invested in. Meanwhile, Bubeck’s former team at Microsoft plans to continue advancing the Phi models, which CEO Satya Nadella has previously described as a key piece of Microsoft’s AI strategy.
The departure also reflects the ongoing mobility within the AI space, where top talent frequently shift between leading tech companies or to start their own initiatives. As the AI landscape evolves, you'll likely see a lot more of these high-profile moves shaping the industry's future.