Alibaba AI Chief Exits After Major Qwen Push
▼ Summary
– Junyang Lin, a central technical leader on Alibaba’s Qwen AI team, announced his abrupt departure from the project just one day after the company unveiled its new Qwen 3.5 small models.
– His departure prompted strong reactions from colleagues and industry partners, who described it as a major loss and credited him with driving the project’s open-source advances and global developer connections.
– The exit occurs amid intense global competition in AI, where Alibaba’s Qwen models are prominent Chinese open-weight efforts that often rival leading U.S. systems in benchmark results.
– Alibaba had just launched its Qwen 3.5 Small Model series, which are native multimodal systems designed for on-device AI and drew attention from figures like Elon Musk for their “impressive intelligence density.”
– The circumstances surrounding Lin’s departure remain unclear, with a project contributor suggesting it was not his choice, and Alibaba did not comment on the reasons or the team’s leadership structure.
A key technical leader for Alibaba’s prominent Qwen artificial intelligence initiative has departed the company, a move that has sent ripples through the project’s team and the broader AI community. Junyang Lin announced his exit just one day after Alibaba unveiled its latest Qwen 3.5 small model series, marking a significant shift for one of China’s most watched open-weight AI efforts. His departure arrives at a pivotal moment of intense global competition, as firms worldwide strive to develop models capable of matching the capabilities of leaders like OpenAI and Google.
Lin revealed his decision to step down in a social media post, offering no specific reasons for the move. He joined Alibaba in 2019 and became an integral part of the Qwen team in April 2023, playing a central role in its technical development and outreach. The timing of his announcement, coming immediately after a major product launch, underscores the project’s rapid pace and the unexpected nature of this leadership change.
The Qwen family of models represents a cornerstone of Alibaba’s AI strategy, having gained recognition for benchmark performances that often compete with leading American systems. The newly released Qwen 3.5 Small Model series includes four variants, designed as native multimodal systems for applications ranging from on-device deployment to lightweight AI agents. The launch garnered notable attention, including public commentary from industry figures like Elon Musk, who praised the models for their “impressive intelligence density.”
Reactions from colleagues and partners highlighted Lin’s profound impact. Wenting Zhao, a Qwen research scientist, described the departure as “the end of an era,” crediting Lin with major contributions to the project’s open-source advances and engineering. Yuchen Jin, CTO of AI infrastructure firm Hyperbolic, noted Lin’s crucial role in bridging connections with the global developer community, recalling collaborative work during critical model launches. Similarly, Tiezhen Wang from Hugging Face’s APAC ecosystem called the loss “immense” for the Qwen project.
The specific circumstances prompting Lin’s exit remain unclear, and he has not publicly elaborated beyond his initial statement. Alibaba has not commented on the leadership change or the future structure of the Qwen team, leaving questions about the project’s immediate direction. Adding to the sense of upheaval, another Qwen team member, Binyuan Hui, has recently updated his social media profile to indicate a past tenure with Alibaba, though it is not confirmed if he has also left the company.
The sentiment within the team was further illustrated by contributor Chen Cheng, who expressed being “heartbroken” by the news. In a poignant social media message seemingly directed at Lin, Cheng wrote, “I know leaving wasn’t your choice,” and mentioned the team had been collaborating on model launches mere hours before the announcement. This sudden transition highlights the human dynamics and potential instability within even the most technically advanced and fast-moving AI development teams as they navigate fierce market competition.
(Source: TechCrunch)
