DeepSeek Core Researcher Daya Guo Rumored to Have Resigned

Daya Guo, a core researcher at Chinese AI firm DeepSeek and a primary author of the landmark DeepSeek-R1 paper that graced the cover of Nature in 2025, has reportedly resigned. Guo, who holds a PhD from Sun Yat-sen University and joined DeepSeek in July 2024, was instrumental in developing key models including DeepSeekMath and DeepSeek-V3. His departure signals a significant loss for the AI unicorn, as he was central to its advanced research on code intelligence and LLM reasoning capabilities.

Industry insiders cite two primary drivers for the move: the lure of vastly superior computing resources at established tech giants and reported internal 'salary inversion,' where newer hires received higher compensation than veteran core members. Guo is rumored to be heading to either Baidu or ByteDance, though neither company has confirmed. This move is seen as part of a broader 'talent war,' with rumors suggesting other key DeepSeek figures are also in talks with major firms seeking greater scope and resources.

The resignation underscores the mounting challenge for well-funded AI startups like DeepSeek in retaining their top technical talent against the immense financial and infrastructural pull of internet behemoths. As the global AI race intensifies, the competition is shifting from pure model development to a fierce battle for the researchers who can push the boundaries of reasoning and efficiency. This trend could reshape the innovation landscape, potentially consolidating cutting-edge research within larger, resource-rich corporations.