Long story short
Alibaba hired a senior DeepMind researcher, Zhou Hao, just as the creator of its Qwen model, Lin Junyang, suddenly resigned along with several key teammates. What appears to be a strategic hire is actually a response to a leadership crisis within Alibaba’s AI division.
What happened
Alibaba had been running an internal system where multiple teams were pushed to compete against each other to build the best AI as fast as possible. Think of it as an internal race with no finish line: fast, productive, but also exhausting.
The system produced strong technical results, including the Qwen 3.5 series, but it also created a culture of pressure, rivalry, and insecurity.
During a leadership reshuffle, Lin Junyang lost influence over the project he had built from scratch. Indeed, he is the person who shaped Qwen’s architecture, training strategy, and early success. He saw decisions being made without him and felt increasingly sidelined. Losing that ownership was a breaking point.
He left, and his exit immediately destabilized the team. Within 48 hours, the leaders responsible for model training and coding also resigned.
This is where DeepMind comes into play. Alibaba rushed to hire Zhou Hao, a respected researcher from Google’s top AI lab, to fill the leadership vacuum and reassure the market that Qwen still had strong technical direction.
This move sends a clear signal to investors and the press: Qwen still has strong technical direction. But it doesn’t solve the real problem: the original core team had already walked out.
The situation escalated so quickly that Alibaba’s CEO had to intervene personally, creating a special task force with the sole purpose of keeping the Qwen project from collapsing entirely.
Remotivate’s take
This is a clear example of what happens when a company prioritizes internal competition over long‑term trust.
You can hire brilliant people from DeepMind, but you cannot replace the deep knowledge, emotional ownership, and continuity that come from the original builders.
When leadership has to step in to contain cultural damage, it shows that the real problem is organizational. In the AI era, your people are the foundation of your strategy, and when they walk out, the technology eventually follows.