Microsoft lost an AI expert and OpenAI gained one with the former's vice president of generative AI research Sebastian Bubeck leaving for the latter. The Information first reported the move, and Reuters also confirmed the company's departure from Microsoft.
We do know that Bubeck is one of the lead authors of a paper describing Microsoft's Phi model, a set of ultra-small language and vision models aimed at pushing AI applications to edge devices. This type of expertise will become more important in some markets as large-scale, centralized models like OpenAI's GPT-4o give way to on-device models that operate quickly, privately, and offline. It's increasing.
Bubeck's new role at OpenAI is still shrouded in mystery, but it seems like he'll likely be working on the efficiency and small-scale model side of things, a corner of the AI world where OpenAI is currently under the radar.