It seems that good news is coming to OpenAI after the last departures from the company of Sam Altman, creator of ChatGPT and Dall-E, among others. According to the company created by Bill Gates, Sebastien Bubeck, one of Microsoft's most important figures, vice president of artificial intelligence at Microsoft, and ten years at the company, is leaving for OpenAI.
The researcher's new role at OpenAI is currently unknown, but a Microsoft spokesperson said the researcher will leave Microsoft to "advance his work toward the development of AGI. " Bubeck's signing is relevant, but in the end, the feeling is that he is staying "in-house" since Microsoft is within OpenAI's investment group, although it is also a rival.
Over the past year, OpenAI has seen an exodus of senior management, including co-founders Ilya Sutskever and Greg Brockman (on sabbatical) and, most recently, CTO Mira Murati. Reports have suggested ideas of prioritizing commercialization over responsible AI development as OpenAI moves further and further away from a research lab and becomes more of a typical for-profit technology company.
Bubeck is vice president of Microsoft GenAI and led a machine learning team in developing Microsoft's Phi models. These lightweight, low-cost versions of OpenAI's GPT models are now used for Bing Chat and various Workplace 365 AI features.
What Bubeck will do in OpenAI is uncertain. However, developing smaller, cheaper AI models is a growing area of focus for all AI companies, as they can perform many day-to-day tasks as larger, basic models in less time and with fewer resources.
OpenAI recently released mini versions of its models, such as GPT-4o mini and o1 mini, a lightweight version of its advanced reasoning model. So, perhaps we will see more profitable models as OpenAI looks to turn a profit and recoup the $6.6 billion investment.
According to Reuters, neither Altman nor Bubeck has made statements about the signing.
