Group 1 - Alibaba announces collaboration with Nvidia for Physical AI [1] - Alibaba officially launches its largest and most powerful model Qwen3-Max to date [2] - Alibaba's Wu Yongming states that AGI is just the starting point, with the ultimate goal being super artificial intelligence [3] Group 2 - The "Stargate" project will establish five AI data centers in the United States [4] - AMD announces expansion of global AI collaboration with AI startup Cohere [5] - SAP and OpenAI announce partnership to launch a "German version of OpenAI" [6] - Applied Materials collaborates with GlobalFoundries to advance AI and AR photonic technologies [7]

当前，通用人工智能已成业界的热点话题，吴泳铭认为，实现AGI，即一个具备人类通用认知能力的智 能系统，现在看来已成为确定性事件。然而，AGI并非AI发展的终点，而是全新的起点。AI不会止步于 AGI，它将迈向超越人类智能、能够自我迭代进化的超级人工智能(ASI)。 他进一步说，AGI的目标是将人类从80%的日常工作中解放出来，从而专注于创造与探索。ASI作为全 面超越人类智能的系统，可能创造出一批"超级科学家"和"全栈超级工程师"，并以难以想象的速度，解 决现在未被解决的科学和工程问题，比如攻克医学难题、发明新材料、解决可持续能源和气候问题，甚 至星际旅行等。 吴泳铭认为，通往ASI有三阶段演进路线：第一阶段是"智能涌现"，AI通过学习海量人类知识具备泛化 智能；第二阶段是"自主行动"，AI掌握工具使用和编程能力以"辅助人"，这是行业当前所处阶段；第三 阶段是"自我迭代"，AI通过连接物理世界并实现自学习，最终实现"超越人"。 中新网杭州9月24日电 (记者 夏宾)云栖大会24日在杭州举行，阿里巴巴集团CEO、阿里云智能集团董事 长兼CEO吴泳铭发表主旨演讲称，最近一年，全球AI(人工智能)行业投资总额已超4 ...

Group 1 - The core viewpoint of the article highlights the significant movement in the Alibaba Cloud concept stocks, particularly the surge in Hangzhou Steel's stock price, which reached the daily limit [1] - On September 24, Alibaba Group's CEO, Wu Yongming, announced at the Cloud Habitat Conference that the energy consumption of Alibaba Cloud's global data centers is expected to increase tenfold by 2032 compared to 2022, indicating a substantial rise in computing power investment [1] - The announcement has led to a rapid increase in the stock prices of related companies such as Hongjing Technology, Zhejiang University Network New, Data Port, and Jinqiao Information [1] Group 2 - The article notes that the developments are in anticipation of the arrival of the ASI (Super Artificial Intelligence) era, which is expected to drive exponential growth in computing capabilities [1] - The mention of 2022 being the "Year of GenAI" sets a context for the expected advancements in technology and infrastructure within the next decade [1]

附吴泳铭演讲全文—— 开始演讲之前，我想特别感谢一下支持整个中国乃至全球科技行业的开发者朋友。今天是云栖大会的10周年，云栖大会起源于阿里云的开发者大会，是广 大开发者推动了中国乃至全球的云计算、AI和科技行业的发展。所以，在演讲之前，我想特别向开发者们致以最高的谢意。 9月24日，阿里巴巴集团CEO吴泳铭在云栖大会上，明确阐述了AI时代下阿里云的核心战略。他判断，"大模型是下一代的操作系统"，将取代传统OS，成 为连接用户、软件与AI计算资源的中间层。基于此判断，吴泳铭宣布阿里云做出关键战略选择：通义千问将坚定走开放路线，致力于成为"AI时代的 Android"，与全球开发者共同构建繁荣的AI应用生态。 吴泳铭解释说，开源模型能渗透的场景和创造的价值将远超闭源模型。目前，通义千问开源模型矩阵的全球下载量已超6亿次，衍生出超过17万个模型， 成为了全球最受欢迎的开源模型之一。通过开源，阿里云旨在将开发者规模从数千万扩展至数亿，让任何人都能用自然语言创造满足自身需求的Agent （智能体），从而"吞噬"传统软件的形态。 要支撑这个全新的"操作系统"和其上无限的应用生态，必须有一台"下一代的计算机"——即"超级A ...

9月24日，以"云智一体·碳硅共生"为主题的2025云栖大会如期在浙江杭州云栖小镇开幕。云栖大会既是 全行业技术演进与产业落地的全景呈现，也历来是阿里巴巴重磅策略发布的舞台。 与往年不同，阿里巴巴集团CEO、阿里云智能集团董事长兼CEO吴泳铭，以长达25分钟的主旨演讲，讲 述了"超级人工智能之路"与阿里的探索与布局。 吴泳铭认为，他认为实现通用人工智能AGI已是确定性事件，但这只是起点，终极目标是发展出能自我 迭代、全面超越人类的超级人工智能ASI。 《中国经营报》记者注意到，阿里巴巴集团董事会主席蔡崇信、吴泳铭等管理层均在不同场分享对AI 发展现状、前景的判断，以及坚持对AI的长期投入。长达25分钟的详尽讲述仍然罕见。 "超级人工智能"的三个阶段 "当前的世界，一场由人工智能驱动的智能化革命刚刚开始。工业革命通过机械化放大了人类技能，信 息革命通过数字化放大了人类的信息处理能力。这一次非常不同，智能化革命将远超我们的想象，通用 人工智能AGI不仅会放大人类智力，还将解放人类迎接超级人工智能ASI的到来。"吴泳铭开场即给出判 断。 在吴泳铭看来，"实现AGI，一个具备人类通用认知能力的智能系统，在行业看来已经 ...

Core Insights - The 2025 Yunqi Conference highlighted the certainty of achieving Artificial General Intelligence (AGI) and the ultimate goal of developing Super Artificial Intelligence (ASI) [1] - Alibaba Cloud's data center energy consumption is projected to increase tenfold by 2032, indicating a significant rise in computational investment [1] Group 1: AI Development Stages - The evolution towards Super AI is outlined in three stages: Emergent Intelligence, Autonomous Action, and Self-Iteration [5] - The current stage is characterized by AI assisting humans through tool usage and programming capabilities [5] - The final stage aims for AI to connect with the physical world and achieve self-learning capabilities [5] Group 2: Strategic Pathways - Alibaba Cloud aims to be a "full-stack AI service provider" with two core strategies: open-source development and the creation of a Super AI Cloud [5] - The open-source model is intended to foster a developer ecosystem and explore limitless AI applications [6] Group 3: Future of AI and Computing - Large models are predicted to become the next generation operating system, allowing users to create applications using natural language [6] - The Super AI Cloud is described as the next generation of computing infrastructure, capable of supporting vast computational needs [10] - It is anticipated that only a few major super cloud computing platforms will exist globally, with AI becoming a critical commodity, surpassing energy in importance [10]

Core Viewpoint - The Hong Kong technology stocks are experiencing a significant rally, driven by strong performances in the semiconductor sector and positive developments from major companies like Alibaba and TSMC [1] Group 1: Market Performance - The Hang Seng Technology Index rose by 1.61%, with Alibaba increasing by 6.65%, reaching a nearly four-year high [1] - Semiconductor stocks such as SMIC and Hua Hong Semiconductor surged over 5%, contributing to the overall market momentum [1] - The Southbound Technology ETF (159269) and Hong Kong Technology ETF (159747) saw increases of 1.69% and 1.5%, respectively [1] Group 2: Company Developments - Alibaba's CEO, Wu Yongming, introduced the concept of "super artificial intelligence" at the 2025 Yunqi Conference and announced plans to invest 380 billion in AI infrastructure [1] - Cathie Wood, known as the "tech goddess," re-entered Alibaba by purchasing approximately 16.3 million USD worth of Alibaba ADRs through her ETFs [1] Group 3: Semiconductor Industry Insights - The semiconductor supply chain is experiencing a boom, with memory chip prices rising and TSMC reportedly increasing prices for its last 3nm CPU process [1] - Shanghai Micro Electronics showcased new equipment, and Changchun Technology's third-quarter report showed significant positive forecasts, boosting market sentiment [1] Group 4: ETF Performance - The Southbound Technology ETF (159269) has seen a net inflow of over 1.16 billion since July 25, with a total size of 1.38 billion, marking a 95.74% increase in just two months [1] - The ETF has the lowest management and custody fees in its category, at 0.3% and 0.05%, respectively [1]

Core Insights - The demand for AI infrastructure is significantly exceeding expectations, prompting Alibaba to invest 380 billion yuan in AI infrastructure development and consider further investments [1] - Alibaba Cloud has launched Qwen3-Max, its largest and most capable model to date, which has achieved a high ranking in text benchmarks, surpassing GPT-5-Chat [1] - The global AI industry is projected to see cumulative investments exceeding $4 trillion over the next five years, marking the largest investment in computing power and R&D in history [4][5] Investment and Market Impact - Following the announcement of Alibaba's AI infrastructure plans, Alibaba's stock price surged over 7% in Hong Kong [2] - Major tech companies, including Nvidia and OpenAI, are also significantly increasing their capital expenditures in AI and infrastructure, with Nvidia planning to invest $100 billion in collaboration with OpenAI [5] Future of AI and Computing Paradigms - The shift from CPU-centric computing to GPU-centric AI computing is revolutionizing data centers, with AI expected to replace energy as a key resource [4] - The concept of "super AI cloud" is emerging, with predictions that only 5-6 super cloud computing platforms will exist globally in the future [4] - AGI (Artificial General Intelligence) is viewed as a new starting point, leading to the development of ASI (Artificial Super Intelligence), which could solve complex scientific and engineering challenges at unprecedented speeds [7][8]

Core Insights - The demand for AI infrastructure is significantly exceeding expectations, prompting Alibaba to invest 380 billion yuan in AI infrastructure development and consider further investments [1] - Alibaba Cloud has launched Qwen3-Max, its largest and most capable model to date, which has achieved a high ranking in text benchmarks, surpassing GPT-5-Chat [1] - The global investment in AI is projected to exceed $4 trillion over the next five years, with a notable shift from CPU-based to GPU-based computing paradigms [3] Investment and Financial Impact - Following the announcement of AI infrastructure investments, Alibaba's stock price surged over 7% in Hong Kong [2] - Recent trends indicate that major tech companies are increasing their capital expenditures in AI and infrastructure [3] Technological Advancements - The next generation of computing, termed "super AI cloud," will require denser computing power, efficient networks, and large-scale clusters, necessitating substantial energy and technology resources [3] - The development of AGI (Artificial General Intelligence) is viewed as a new starting point, leading towards ASI (Artificial Super Intelligence), which could revolutionize various fields [5][6] Future Projections - The rapid consumption of AI tokens and the significant increase in global AI investments highlight the accelerating pace of AI development [3] - The establishment of large-scale data centers by companies like NVIDIA and OpenAI indicates a strong commitment to building the necessary computational infrastructure for AI advancements [4]

2025.09.24 本文字数：1015，阅读时长大约2分钟 作者 |第一财经 陈杨园 "AGI并非人工智能发展的终点，而是起点，它将迈向超越人类智能、能够自我迭代的超级人工智能 （ASI）。"9月24日，2025云栖大会开幕式上，阿里巴巴集团CEO吴泳铭发表主题演讲《超级人工智能 之路》，并作出了上述论断。 吴泳铭表示，过去三年，阿里看到的AI演进路线越来越清晰，AI的行业渗透率将超过历史上所有的技 术，阿里看到，token的消耗每两三个月会翻一番，AGI会具备人类通用认知能力在行业看来已经是确定 性的事情。 AGI并不是终点，吴泳铭认为，AI会经历三个阶段最终成长为超级人工智能。第一阶段是智能涌现，AI 学习人；第二阶段是AI自主行动，辅助人，我们刚刚处在这个阶段的开端，未来也许会有超过世界人 口的智能体和机器人和人类一起工作；第三个阶段是自我迭代，超越人，跨越到这个阶段需要两个要 素，AI将逐步连接几乎物理世界的所有场景和数据，模型能够自我学习、通过与真实世界的持续交互 获得新的数据实现自我迭代与智能升级。 吴泳铭相信，未来大模型将运行在所有计算设备中，基于此，阿里巴巴坚持开源路线，目标是打造AI 时代 ...