<img src="https://imageproxy.pbkrs.com/https://wpimg-wscn.awtmt.com/73f205fe-0c3d-4144-b81c-992acac5d593.png?x-oss-process=image/auto-orient,1/interlace,1/resize,w_1440,h_1440/quality,q_95/format,jpg" alt="" original-src="https://imageproxy.pbkrs.com/https://wpimg-wscn.awtmt.com/73f205fe-0c3d-4144-b81c-992acac5d593.png"/>
Author | Huang Yu
When Alibaba's open-source large model frequently topped the global developer community and even received a thumbs-up from Musk, an unexpected personnel shock quietly descended in the middle of the night.
In the early hours of March 4, Lin Junyang, the &#34;soul figure&#34; of Alibaba's Qwen large model, unexpectedly announced his resignation on social platform X. This news instantly ignited the global AI circle. Less than 48 hours prior, he had just led his team to release the Qwen 3.5 small model series, garnering top attention worldwide.
A widely circulated explanation for Lin Junyang's resignation is that Alibaba Cloud's senior management made structural adjustments to the Qwen team, parachuting in a new technical leader, which may have changed Lin Junyang's influence, leading to disagreements in direction and ultimately his departure.
As of the time of publication, Alibaba has not responded to this matter.
The departure of the soul figure has led the outside world to begin examining the stability of Alibaba's AI talent pipeline. However, just a day before Lin Junyang's announcement, Jack Ma, who had not made a public appearance for a long time, led the entire core management team of Alibaba and Ant Group to gather at the Hangzhou Yun Gu School, founded by Alibaba partners, to publicly discuss the challenges and opportunities of the AI era.
On one side is the &#34;sudden&#34; departure of a core technical figure, and on the other side is the founder and top decision-making team collectively supporting AI. This seemingly contradictory scene reflects Alibaba's new AI blueprint, which is currently in a strategic transition period, full of uncertainties yet resolute.
<h2>AI Team Earthquake</h2>
In today's AI landscape, the name Lin Junyang represents an extreme &#34;technical myth.&#34;
As Alibaba's youngest P10-level technical leader, born in 1993, he is only 32 years old this year, yet he has already been a core driver of Alibaba's Qwen series of open-source large models.
In the early hours of March 4, Lin Junyang suddenly posted on social media: &#34;me stepping down. bye my beloved qwen.&#34;
Although he did not explicitly state whether he was leaving Alibaba entirely, &#34;leaving the Qwen team&#34; alone was enough to shake the industry.
After the announcement, many AI researchers close to Qwen posted specific emojis in their social circles to show respect, and some researchers even raised the slogan &#34;Qwen is nothing without its people.&#34;
Chen Cheng, who is described as a maintainer of the Modao community and a contributor to Qwen, commented that he was heartbroken, believing that the departure was not Lin Junyang's own choice. The night before (on March 2), they had jointly released the Qwen 3.5 small model series, and he could not imagine Qwen without Lin Junyang.
This emotional expression evokes memories of the scene when Sam Altman left OpenAI
The reason this change seems extremely sudden is that just on the evening of March 2, Lin Junyang led his team to open-source four Qwen3.5 small-sized models, quickly attracting attention from the overseas tech community, with even Elon Musk commenting under the tweet, &#34;Amazing intelligence density.&#34;
Announcing his departure at the peak of glory has sparked countless speculations about the reasons behind it, but there has been no official confirmation of his exit so far.
Lin Junyang's departure is undoubtedly seen as having an impact on the development of the Qianwen large model in the industry.
Lin Junyang has a cross-disciplinary educational background in computer science and linguistics, which has laid a unique advantage for his deep engagement in multimodal large models. After graduating with a master's degree in 2019, he joined Alibaba, focusing on the research and development of AI underlying models, quickly rising from a grassroots engineer to Alibaba's youngest P10-level technical expert.
At the end of 2022, as the wave of AI large models surged, Alibaba merged the AI teams in language, vision, and other areas from Damo Academy into Alibaba Cloud, establishing the Tongyi Laboratory, led by Alibaba Cloud CTO Zhou Jingren. Lin Junyang was officially appointed as the technical head of the Tongyi Qianwen series large models under the Tongyi Laboratory, fully responsible for core model research and strategic planning, becoming a key figure in Alibaba's large model track.
Lin Junyang led the open-source efforts for the Qianwen series large models, positioning Alibaba Qianwen among the top global open-source large models with significant global influence.
The Stanford University Artificial Intelligence Index Report 2025 pointed out that the performance gap between top AI large models in China and the U.S. has narrowed significantly to 0.3%, nearly closing, with Alibaba's Qwen contributing the third-highest globally.
Additionally, Jensen Huang, CEO of NVIDIA, revealed during the 2025 GTC conference that Qwen has the highest market share among global open-source models and is still growing.
The official MiniMax account also commented on Lin Junyang's announcement of his resignation, saying, &#34;Thank you for your contributions to the open-source community.&#34;
At the AGI-Next Frontier Summit in January this year, Lin Junyang participated in a roundtable discussion as the technical head of Alibaba Qianwen, alongside Tang Jie, founder of Zhipu AI, and Yao Shunyu, chief AI scientist at Tencent.
Regarding the future of AI, Lin Junyang believes, &#34;If your idea is not to help all of humanity, then it's better not to create a large model.&#34; He hopes that future models will not only be top students who pass exams but also true agents that can genuinely help humanity. He firmly believes that embodied AI, which can operate computers, write code, and even serve tea in the physical world, is the ultimate form of AI transitioning into the real world.
Thanks to the achievements and reputation gained by Alibaba Qianwen in recent years, even if Lin Junyang leaves Alibaba, he will clearly not lack opportunities.
It is worth mentioning that Lin Junyang's departure is not an isolated case; since 2024, several technical executives from Alibaba's large model team have left one after another, including Huang Fei, head of natural language processing (NLP) at Tongyi Laboratory, Bo Liefeng, former head of the application vision team, Yan Zhijie, former head of the voice team, and Zhou Chang, who joined ByteDance to oversee AI business.
According to reports, on the same day Lin Junyang announced his resignation, Yu Bowen, head of post-training for Qianwen, also officially left, with his successor being Zhou Hao, a former senior researcher at DeepMind who joined Tongyi Laboratory earlier this year
Qwen Code head Hui Bin also left Alibaba in January 2026 to join Meta.
In the current context of fierce competition for top AI talent, the high-frequency movement of such elite personnel seems to have become the norm.
<h2>Where is Alibaba's AI Heading?</h2>
Just a day before Lin Junyang announced his resignation, on March 3rd, the core management team of Alibaba Group and Ant Group rarely gathered together at Hangzhou Yun Valley School.
This appearance was of high significance: Jack Ma, Alibaba Group Chairman Joseph Tsai, CEO Eric Wu, Risk Committee Chairman Shao Xiaofeng, e-commerce business group CEO Jiang Fan, Ant Group Chairman Jing Xiandong, and CEO Han Xinyi were all present.
This &#34;all-star lineup&#34; gathering seems to convey a clear message to the outside world: even with some personnel changes, Alibaba's strategic investment in AI will not waver.
During the discussion, Jack Ma candidly stated that the AI era has arrived rapidly, and its impact on society will exceed imagination, even suggesting that &#34;none of us are adequately prepared.&#34;
Currently, AI iterations are measured in weeks, and capabilities are continuously growing. This technological revolution brings historic changes to productivity and various aspects of society; in the future, it may be possible to work less than 8 hours a day, leading to immense societal wealth, but many job types will disappear. Jack Ma stated: &#34;The reason we all came to Yun Valley is to tell everyone that this change will come very quickly, and we must make rapid changes to help children learn to coexist with AI from now on and adapt to this enormous change.&#34;
Although the turnover of core technology leaders brings uncertainty, Alibaba's confidence in AI comes from its established &#34;golden triangle.&#34;
Recently, the &#34;Tongyun Ge&#34; architecture, composed of Tongyi Laboratory, Alibaba Cloud, and Pingtouge, has emerged for the first time, indicating that Alibaba has full-stack self-developed chips, robust cloud infrastructure, and globally leading open-source models, making it one of only two tech companies in the world with top-tier capabilities in large models, cloud, and chips, the other being Google.
This Spring Festival, Alibaba also vigorously promoted the C-end-oriented Qianwen APP with a &#34;3 billion yuan Spring Festival banquet plan,&#34; offering 25 yuan no-threshold coupons covering diverse scenarios such as takeout, instant shopping, movie tickets, flights, and hotels, and extended the coupon validity multiple times until March 3rd.
Morgan Stanley estimates that Alibaba's order subsidy alone reached 5 billion yuan (approximately 200 million orders × 25 yuan), and with brand marketing expenses added, the actual expenditure exceeded 5 billion yuan. In terms of investment intensity, Alibaba's input far exceeds that of Tencent, ByteDance, etc. The DAU of Qianwen APP has seen a significant increase, but user retention remains a challenge.
Regardless, Alibaba's determination to promote the Qianwen APP is beyond doubt.
Alibaba's AI strategy will not undergo drastic swings due to individual departures. For Alibaba, this technological revolution is not only a competition for productivity but also a reshaping of the logic of future survival

09988

BABA

YINN

KWEB

BABX

KBAB

BABO

- Alibaba's AI team faced a significant leadership shake-up with Lin Junyang's unexpected resignation as the key figure behind its Qwen models, announced on March 4.  
- Despite this shift, Alibaba's commitment to AI remains steadfast, as highlighted by founder Jack Ma's recent gathering with top management to discuss the challenges and opportunities presented by AI.  
- The departure of Lin, a prominent tech leader, raises questions about the stability of Alibaba's AI talent pool, yet the company's strategic investments and innovations signal a resilient future in AI development.

Wallstreetcn

BABA-W

Alibaba

Direxion FTSE China Bull 3X

Krne Csi China Internet

Invesco Golden Dragon China

BABA 2x Long Daily ETF - GraniteShares

KraneShares 2x Long BABA Daily ETF

YieldMax BABA Option Income Strategy ETF

On the eve of the departure of the "soul figure" of Qianwen, Jack Ma leads the management team to release signals regarding the AI strategy