据南方都市报报道，今天（3月4日）凌晨，阿里Qwen技术负责人林俊旸在社交平台上与千问 道别，称"我退出了"。 有业内人士猜测， 林俊旸的离开或由于其考核目标变成了日活跃用户数。 对于林俊旸离开后 谁 会 接 任 其 位 置 ， 该 业 内 人 士 也 提 到 或 是 一 位 来 自 谷 歌 Gemini 团 队 、 以 日 活 跃 用 户 数 （DAU）指标为导向的非核心人才。南都记者就此向阿里云方面求证，截至发稿尚无回复。 林俊旸与月之暗面创始人兼CEO杨植麟、腾讯首席AI科学家姚顺雨和清华大学教授、智谱创 始人兼首席科学家唐杰被在业内并称"基模四杰"， 其他几位均为月之暗面、腾讯、智谱等企 业的模型掌门人。 今年1月在中关村，林俊旸还提到，自己2021年做模型的时候，阿里芯片团队的人来问自己三 年后模型主流架构是不是Transformer和多模态，因为三年后模型才能流片。他表示，"三年之 后 ， 我 在 不 在 阿 里 巴 巴 ， 我 都 不 知 道 。 但 最 后 我 今 天 还 在 阿 里 巴 巴 ， 然 后 他 果 然 还 是 Transformer，还是多模态，我就非常懊悔为什么当时没有去催他去 ...

Group 1 - The core focus of the article is the launch of Meituan's 2026 spring campus recruitment, aiming to hire 6,000 graduates from the 2026 class and over 3,000 internship positions for the 2027 class [1] - The recruitment will cover 10 major categories and over 100 types of positions, including technology, product, business analysis, operations, finance, retail, functions, marketing, design, sales, customer service, and support [1] - The recruitment will take place in various global cities, including Beijing, Shanghai, Shenzhen, Chengdu, Guangzhou, Hong Kong, and Dubai [1] Group 2 - The "Beidou Plan" is a key initiative aimed at recruiting top campus technology talent throughout the year, focusing on cutting-edge technology areas such as large model foundations, large model applications, autonomous driving, drones, and intelligent decision-making [1]

Core Viewpoint - The sudden resignation of Lin Junyang, a key figure in Alibaba's AI development, marks a significant transition for the Qwen project and raises concerns about its future direction and management [2][5][7]. Group 1: Lin Junyang's Background and Contributions - Lin Junyang, born in 1993, was the youngest P10-level technical expert at Alibaba, with a strong academic background in computer science and linguistics from Peking University [2]. - He joined Alibaba's DAMO Academy in 2019 and became a significant contributor to the M6 model team, which was established as the foundation for Alibaba's general large model technology [3][4]. - Under his leadership, the Qwen series of open-source models was launched, overcoming critical challenges in reinforcement learning and multilingual capabilities [4]. Group 2: Organizational Changes and Future Implications - Following Lin's departure, there are speculations about a shift in Qwen's management structure, with reports of increased oversight from Alibaba Cloud's CEO and the introduction of a new leader from the Gemini team [5]. - The potential transition of Qwen models towards a closed-source approach has been suggested, indicating a possible shift in Alibaba's strategy to align more closely with commercial models seen in competitors like Google Cloud [5]. Group 3: Industry Reactions and Community Impact - Lin's resignation has been met with widespread acknowledgment in the industry, with peers recognizing his contributions to the open-source community and the advancement of AI technologies [7][12]. - Comments from industry figures highlight the importance of leadership in fostering community engagement and innovation, emphasizing that the mission and vision of leaders are crucial for the future of projects like Qwen [12][13].

编辑 | 七三 3月3日夜间，阿里千问大模型技术负责人林俊旸自宣卸任，他在社交平台写道"me stepping down. bye my beloved qwen."（我将卸任，再见我亲爱的千问）。作为阿里千问大模型的技术核 心人物，林俊旸或将离开千问团队，目前尚不知他是否会离开阿里巴巴。记者就此消息向阿里巴巴求 证，截至发稿暂未获得回复。 ...

Core Viewpoint - The article discusses the current state of three domestic AI chip companies: Cambricon, Muxi, and Moore Threads, highlighting their financial performance and market expectations amidst a booming AI computing market valued at 600 billion [2][4]. Group 1: Financial Performance - Cambricon reported a total revenue of 6.497 billion, a year-on-year increase of 453.21%, and a net profit of 2.059 billion, marking its first profit since going public [5][6]. - Muxi achieved a revenue of 1.644 billion, with a year-on-year growth of 121.26%, but incurred a net loss of 778 million, although this loss was reduced by 44.53% [5]. - Moore Threads generated a revenue of 1.506 billion, reflecting a year-on-year increase of 243.37%, but faced a net loss of 1.024 billion, with a reduction in losses by 36.7% [5]. Group 2: Market Expectations and Challenges - Despite Cambricon's profitability, the market reaction was lukewarm, as analysts noted that the results fell short of expectations, with market sentiment not fully aligning with the company's performance [4][9]. - Analysts expressed concerns about Cambricon's growth sustainability, citing a decline in profit margins and potential supply chain issues, which could hinder future performance [10][20]. - The emergence of self-developed chips by major clients like ByteDance and Alibaba poses a competitive threat to Cambricon, leading to a decline in its stock price [14][15]. Group 3: Opportunities for Muxi and Moore Threads - Muxi and Moore Threads are positioned to potentially benefit from the demand gap left by Cambricon, but their current market penetration remains limited, primarily serving secondary scenarios and relying on government projects [21][22]. - Both companies are seen as needing to secure significant contracts with cloud service providers to ensure sustainable revenue growth and operational cash flow [23][24]. - Muxi has focused on optimizing its sales strategy towards larger projects, while Moore Threads is leveraging its geographical advantage to secure more government contracts [24]. Group 4: Industry Dynamics and Future Outlook - The AI chip industry is facing new challenges, including the rise of new players and the potential onset of a price war, which could impact profit margins across the sector [25][26]. - The competition is shifting from single-card performance to a focus on comprehensive capabilities, including memory, bandwidth, and system software integration [26]. - The future landscape of domestic AI chips will depend on production capacity, delivery capabilities, and the ability to adapt to market demands [26].

Core Viewpoint - The sudden resignation of Lin Junyang, the technical head of the Qwen large model, has sparked significant discussion within the industry, raising questions about the future of the Qwen project and the team dynamics at Alibaba [1][4]. Group 1: Lin Junyang's Background - Lin Junyang, born in 1993, is noted as Alibaba's youngest P10-level technical executive, having joined Alibaba's Damo Academy in 2019 after graduating from Peking University with a master's degree [2]. - He was promoted to P9 in August 2024 and then to P10 in May 2025, indicating a rapid career progression within the company [2]. Group 2: Team Reactions and Implications - The announcement of Lin's departure has led to over 1000 comments on social media, with team members expressing deep disappointment and concern about the future of the Qwen project without his leadership [3]. - Multiple core contributors to the Qwen team have also confirmed their departure, suggesting potential instability within the team and raising concerns about the project's continuity [3]. Group 3: Uncertainty Surrounding Qwen Team - The specific reasons behind the departures within the Qwen large model team have not been disclosed, leaving uncertainty about the internal issues and whether Alibaba will provide further clarification [4].

Investment Rating - The report maintains a "Positive" investment rating for the software and services industry [8]. Core Insights - Recent explosive growth in AI model invocation in China, with OpenRouter platform's invocation surpassing that of the US for the first time [2][5]. - The rapid increase in invocation volume indicates a strong collective rise of Chinese AI manufacturers, not reliant on a single blockbuster product [10]. - The report anticipates accelerated commercialization and international expansion of domestic AI models, driven by significant price competitiveness [10]. Summary by Sections Industry Overview - China's AI model invocation reached 41.2 trillion tokens, exceeding the US's 29.4 trillion tokens for the week of February 9-15, 2026 [10]. - The following week saw further growth to 51.6 trillion tokens, marking a 127% increase over three weeks, while US invocation dropped to 27 trillion tokens [10]. Model Performance - In the week of February 16-22, 2026, four out of the top five models by invocation volume were from Chinese manufacturers, contributing to 85.7% of the total [10]. - Notable models include MiniMax's M2.5, which contributed 14.4 trillion tokens, and KimiK2.5, which saw its revenue exceed its total for the previous year shortly after launch [10]. Competitive Landscape - Price competitiveness remains significant, with MiniMax-M2.5 and GLM-5 priced at $0.3 per million tokens, compared to Claude Opus4.6 at $5, highlighting a 16.7 times difference [10]. - The report suggests that the value of tokens is shifting from internet traffic to essential resources in the AI era, indicating a potential new business model [10]. Future Outlook - The year 2026 is projected to be a pivotal year for the commercialization of large models, with key changes anticipated in business models and market dynamics [10]. - The report emphasizes three main themes: new entry points and commercialization of large models, domestic chip development, and the restructuring of software through agents [10].

上海重点产业招聘需求旺盛 上海该场春季招聘会以"春风送岗促就业 精准服务助发展"为主题，聚焦"3+6+2"重点行业领域和高校毕业生等重点群体就业需求，共有1000家企 业参加，提供超过1.3万个招聘岗位。 当天上午，记者到达招聘会展馆时，应聘者正在工作人员的引导下，根据预约时间段有序进场，入口前的招聘企业名录吸引了不少应聘者驻足。 人工智能（AI）等产业揽才势头足！ 3月3日，2026年上海市春季促进就业专项行动暨高校毕业生择业对接会在上海世贸商城举行，这是马年春节后上海首场大型招聘会。证券时报记 者实探发现，AI等重点产业企业招聘需求依然旺盛，其中大模型及相关算法类岗位需求持续扩容。 上海市就业促进中心主任周国良接受记者采访时表示，今年用人单位招聘积极性比往年要高，在招聘会筹备的几天时间内，愿意到线下来招聘的 用人单位就达到约1300家，但招聘场地只能容纳1000家，于是就引导多家企业参与线上招聘。 AI相关岗位薪酬依然最高 人工智能展区依然是上海此次招聘会最热的专区，招聘通道被应聘者围得水泄不通。多家科技类招聘企业，只能见缝插针地与应聘者聊上几句。 记者在现场看到，此次招聘会共设置17个招聘专区，包含人工 ...

Investment Rating - The report maintains a positive investment rating for leading internet companies with strong performance and AI sector positioning, such as Tencent Holdings, Kuaishou, Bilibili, and Meitu [1]. Core Insights - The internet industry is witnessing accelerated iterations in applications and edge-side technologies, with major companies set to release their 2025 annual reports [1]. - Keep is expected to narrow its losses significantly in 2025, projecting a loss of approximately RMB 72 million, down 87% from RMB 534.7 million in 2024, and is anticipated to achieve an adjusted profit of RMB 25 million [2]. - The gaming company Ganwan has signed agreements to acquire a 35% stake in Shanghai Yuelong and gain exclusive operational rights for the game "Tianlong Babu 2: Feilong Zhantian" [3]. Industry Overview - The internet industry comprises 160 listed companies, with a total market capitalization of RMB 2,087.1 billion and a circulating market capitalization of RMB 1,920.8 billion [5]. - The absolute performance of the industry over the past month is -6.1%, with a 6-month performance of 6.9% and a 12-month performance of 19.7% [7]. Company Performance - In the past week, Keep's stock increased by 2.87%, while Ganwan's stock decreased by 14.82% [22]. - Tencent Holdings has a current market value of RMB 47,170.92 million, with a weekly decline of 0.77% [25]. - Kuaishou's market value is RMB 2,735.46 million, with a weekly decline of 5.49% [26]. Recommended Companies - The report highlights key companies to watch, including Tencent Holdings, Kuaishou, Bilibili, and Meitu, due to their strong positioning in the AI sector and overall performance [9].

Core Viewpoint - The article emphasizes that the AI competition has shifted from a parameter race to a focus on system efficiency and business viability, as evidenced by MiniMax's 2025 financial results, marking the entry into a new phase of AI competition [2]. Group 1: Financial Performance and Business Model Validation - MiniMax reported total revenue of $79.038 million in 2025, a year-on-year increase of 158.9%, driven by AI native product revenue of $53.075 million (up 143.4%) and open platform revenue of $25.963 million (up 197.8%) [3]. - Over 73% of MiniMax's revenue came from international markets, indicating successful global market penetration [3]. - The company achieved a significant reduction in adjusted net loss rate, demonstrating a balance between scale expansion and commercial efficiency [4]. Group 2: Operational Efficiency and Cost Management - R&D expenses increased by 33.8% to $253 million, while total revenue growth outpaced this increase, indicating the onset of operational leverage [5]. - Gross margin improved from 12.2% in 2024 to 25.4% in 2025, attributed to enhanced model and system efficiency, particularly in reducing inference costs [5]. - Sales and distribution expenses decreased by 40.3% to $51.90 million, highlighting strong organic growth driven by product appeal [5]. Group 3: Acceleration of Evolution and Competitive Advantage - The speed of model iteration has become a key competitive metric, with MiniMax releasing multiple models within a compressed timeline, showcasing rapid capability enhancement [7][8]. - The M2 series models demonstrated a significant increase in daily token consumption, indicating strong market adoption and performance [12]. - MiniMax's internal AI systems, such as the Forge framework, have accelerated training processes by approximately 40 times, enhancing overall organizational efficiency [14][15]. Group 4: Future Directions and Market Trends - The emergence of the Agent era is highlighted, with MiniMax's M2 model achieving significant recognition in developer communities and surpassing 50 billion daily token consumption on OpenRouter [17]. - The demand for AI is shifting from passive interactions to active task execution, suggesting a future increase in token consumption for complex tasks [23]. - MiniMax aims to transition from a large model company to an AI platform company, focusing on intelligent density and model throughput as core metrics [25][26].