I just noticed an exciting development in the Chinese artificial intelligence world - a sudden resignation that shook the entire tech scene. Lin Junyan, the man who led Qwen to the top of global open-language models, left his position at Alibaba in a dramatic manner.

The story is more complicated than it appears. On the morning of March 4th, Lin Junyan posted a simple message on the X platform: "me stepping down. bye my beloved qwen." - and in an instant, Alibaba lost its technical leader. But there is something deeper than that. According to multiple reliable sources, this was not a voluntary resignation but the result of intense disagreements within the organization.

What happened immediately before that? On the night of March 2nd, Lin Junyan’s team launched four small models of Qwen3.5, receiving praise from Elon Musk himself, who commented "Incredible intelligence density." On the same day, Alibaba announced the unification of its large model strategy under the "Qianwen" brand. Then, in the evening, Lin Junyan left an internal meeting directly due to disagreements and submitted his resignation. Some colleagues cried upon hearing the news.

The real reason? It relates to a fundamental disagreement over the vision for model development. Lin Junyan believes that true power comes from integrated collaboration across all team divisions — pre-training, fine-tuning, multimedia development, all together. But Alibaba wants to restructure the team horizontally, dividing it into separate units. This means significantly reducing Lin Junyan’s authority.

There is also a deeper contradiction — between open-source strategy and profit goals. Under his leadership, Qwen became a global standard in open models. On Hugging Face, models derived from Qwen exceeded 200,000 models with over a billion downloads. But Alibaba has started to feel that open-source isn’t generating enough profit. Some managers described Qwen3.5 as a "deficient product," and commercial pressure has begun to weigh heavily on the technical side.

The organizational landscape has also changed. Alibaba brought in global experts — including former researchers from DeepMind — placing them under direct supervision. The lab shifted from a single-core model led by Lin Junyan to a "multi-force parallel" structure. This gradually weakened his position.

The impact is now clear — in just three months, the core team founders resigned: the head of technology, the head of fine-tuning, and the head of coding. Each of them left Alibaba. It is said that senior executives are still trying to persuade Lin Junyan to return, but the likelihood is very low. Most probably, he will start his own project or join a star team in the model field.

This moment marks a real turning point. Alibaba is bidding farewell to an era focused on building a global technical standard and is shifting entirely toward a cycle centered on commercial transformation. But the question remains: can Alibaba make this transition without losing the technical momentum it has built? The disruption caused by the loss of the core team, combined with fierce competition from ByteDance and Tencent, will truly test Alibaba’s strategic resilience in the field of artificial intelligence.