通用人工智能(AGI)
Search documents
阿里最新架构变动!
证券时报· 2025-12-10 00:11
证券时报·券商中国记者日前获悉,阿里已成立千问C端事业群,由阿里巴巴集团副总裁吴嘉负责。 据悉,该事业群由原智能信息与智能互联两个事业群合并重组而来,包含千问APP、夸克、AI硬件、UC、书 旗等业务。 这也是今年9月宣布的额外AI基础设施投入的一部分。今年9月,吴泳铭概述了他自己推出新模型和"全栈"AI 技术的计划,这反映了阿里巴巴既要开发服务,也要开发支撑该技术的基础设施的意图。 9月24日,阿里巴巴集团CEO、阿里云智能集团董事长兼CEO吴泳铭在云栖大会演讲中表示,大模型是下一代 操作系统,而AI云是下一代计算机。也许未来全世界只会有五六个超级云计算平台。目前阿里正积极推进 3800亿元的AI基础设施建设,并计划追加更大的投入。 吴泳铭认为,实现AGI(通用人工智能)已是确定性事件,但这仅是起点,终极目标是发展出能自我迭代、 全面超越人类的ASI(超级人工智能),以解决气候、能源、星际旅行等重大科学难题。 通往超级人工智能之路分为三个阶段:一是"智能涌现",AI通过学习人类知识具备泛化智能;二是"自主行 动",AI掌握工具使用和编程能力以"辅助人",这是行业当前所处的阶段;三是"自我迭代",AI通过连接 ...
梁文锋,Nature全球年度十大科学人物!
具身智能之心· 2025-12-10 00:03
Core Viewpoint - Liang Wenfeng has been recognized as one of the top ten scientists of 2025 by the journal Nature for his significant contributions to the AI field through the DeepSeek model, which has disrupted traditional AI paradigms [3][4][8]. Group 1: DeepSeek Model and Its Impact - The DeepSeek model has dramatically reduced costs in the AI industry while enhancing the global visibility of domestic large models [9]. - DeepSeek demonstrates that high-performance models do not necessarily require vast amounts of data, parameters, or servers to achieve top-tier capabilities [10]. - The recent release of the DeepSeek V3.2 series model has achieved the highest evaluation level among current open-source models in the Agent domain [11][12]. Group 2: Liang Wenfeng's Background - Liang Wenfeng, born in 1985, excelled academically, becoming a top student in his high school and later studying electronic information engineering at Zhejiang University [14][15]. - He transitioned into quantitative investment in 2008, capitalizing on the emerging trend of quantitative trading in China, and his team quickly grew their proprietary funds to over 500 million yuan [17]. - In 2021, his firm, Huanfang Quantitative, became one of the largest quantitative private equity firms in China, managing over 100 billion yuan [19]. Group 3: Recognition of Other Researchers - Mengran Du, another researcher recognized by Nature, discovered the deepest known animal ecosystem on Earth, contributing to the understanding of extreme life and carbon cycling in deep-sea environments [25][28]. - Du's research has been published in prestigious journals and she has participated in numerous deep-sea explorations, enhancing the scientific community's knowledge of deep-sea ecosystems [33].
被OpenAI开除的年轻人投资了一家为OpenAI对手提供算力的初创
Sou Hu Cai Jing· 2025-12-09 10:06
一个被 OpenAI 扫地出门的 23 岁德国年轻人,如今正准备领投一家专门为 OpenAI 竞争对手提供算力的初创公司,谷歌也准备入局,甚至计划借此大规模 部署自研的 TPU。这笔交易若能达成,Sam Altman 大概不会太高兴。 X) 图丨Leopold Aschenbrenner(来源: 根据彭博社 12 月 4 日的报道,云计算初创公司 Fluidstack 正在洽谈一轮约 7 亿美元的融资,完成后估值将达到 70 亿美元。领投方是前 OpenAI 研究员 Leopold Aschenbrenner 创立的对冲基金 Situational Awareness LP(简称 SA)。Google 也在讨论参与,高盛担任财务顾问。Fluidstack、Google 和高盛均拒 绝置评,Aschenbrenner 本人也未回应。 只是这笔交易有意思的地方不在于金额,70 亿美元的投资在当下的数据中心热潮可能已不足为奇,但这家公司与领投方复杂的背景,却大有文章。 Fluidstack 是当下最火的"新云"(Neocloud,指专门为 AI 公司提供高性能 GPU 算力的新型云计算企业)公司之一。Fluidst ...
云知声CEO黄伟斩获2025年度“安永企业家奖”殊荣
Sou Hu Cai Jing· 2025-12-09 09:29
Core Insights - The article highlights the recognition of Huang Wei, CEO of Yunzhisheng, as the winner of the "Ernst & Young Entrepreneur Award 2025" for his contributions to the commercialization of General Artificial Intelligence (AGI) and his leadership in the tech industry [1][11] - The award marks the 20th anniversary of the event and symbolizes a collective tribute to a new generation of Chinese tech entrepreneurs who empower industrial upgrades through innovative productivity [1] Company Overview - Yunzhisheng, founded in 2012, is one of the pioneers in the research and industrial application of AI voice and semantic technologies in China [4] - The company has made significant breakthroughs in key sectors such as healthcare, home automation, transportation, and government services, establishing a technology-centric industrial ecosystem [5] Technological Advancements - Huang Wei has strategically positioned Yunzhisheng in the "professional-grade large model" sector, with its self-developed "Shanhai Model" demonstrating leading capabilities in various global assessments [7] - The company has achieved top rankings in general capability evaluations and specialized medical assessments, showcasing its technical versatility and domain adaptability [7] Market Recognition - Yunzhisheng went public on the Hong Kong Stock Exchange in June 2025, becoming the "first stock of AGI," which reflects strong market recognition of its sustainable development path [7] Entrepreneurial Philosophy - Huang Wei emphasizes the importance of "reconstructing competitiveness" in the rapidly evolving AI landscape, advocating for a comprehensive capability that includes rapid learning, foresight, and agile response [10] - The company embodies a dual approach of "inheritance + innovation," maintaining core technologies and corporate culture while pushing the boundaries of technology and application [10] Industry Context - Huang Wei views the "bubble" in the AI sector as a catalyst for technological transformation, attracting resources and talent towards hard-core technology, ultimately driving industrial upgrades [10] - The article reflects on the broader narrative of China's AI industry growth, with Yunzhisheng's journey symbolizing the resilience and adaptability of new-generation tech entrepreneurs [11]
AI需要能自我改进!AI圈越来越多人认为“当前AI训练方法无法突破”
Hua Er Jie Jian Wen· 2025-12-09 01:49
来自OpenAI、谷歌等公司的小部分但日益增长的AI开发者群体认为,当前的技术路径无法实现生物 学、医学等领域的重大突破,也难以避免简单错误。这一观点正在引发行业对数十亿美元投资方向的质 疑。 据The Information周二报道,上周在圣地亚哥举行的神经信息处理系统大会(NeurIPS)上,众多研究 人员讨论了这一话题。他们认为,开发者必须创造出能在部署后持续获取新能力的AI,这种"持续学 习"能力类似人类的学习方式,但目前尚未在AI领域实现。 然而,技术局限已拖慢企业客户对AI代理等新产品的采购。模型在简单问题上持续犯错,AI代理在缺 乏AI提供商大量工作确保其正确运行的情况下往往表现不佳。 这些质疑声与部分AI领袖的乐观预测形成对比。Anthropic首席执行官Dario Amodei上周表示,扩展现有 训练技术就能实现通用人工智能(AGI),OpenAI首席执行官Sam Altman则认为两年多后AI将能自我 改进。但如果质疑者是对的,这可能令OpenAI和Anthropic明年在强化学习等技术上投入的数十亿美元 面临风险。 尽管存在技术局限,当前AI在写作、设计、购物和数据分析等任务上的表现仍推 ...
Nature年度十大人物发布:DeepSeek梁文锋、中科院深海所杜梦然入选
生物世界· 2025-12-09 00:05
撰文丨王聪 编辑丨王多鱼 Nature 特写部编辑 Brendan Maher 表示: 今年的榜单颂扬了对新前沿的探索,医疗领域突破性进 展的希望,对科研诚信的坚定守护,以及那些可以拯救生命的全球政策的制定者。我们激动地看到这么多人在不遗余力地理解自然世界,并在许多情况下帮助 排版丨水成文 | 日前, | | Nature | | 公布了 | 2025 | 年度 | 十大人物榜单 | (Nature's 10) | | --- | --- | --- | --- | --- | --- | --- | --- | --- | | ,入选者包括 | | | DeepSeek 创始人 | | | 梁文锋 | 、 | 中国科学院深海科学与工 | | 程研究所 | 杜梦然 | | 研究员 | | ,以及 | Susan Monarez | 、 | Achal Agrawal | | 、 | Tony Tyson | | 、 | | Precious Matsoso | 、 | Sarah Tabrizi | 、 | | Luciano Moreira | | 、 | Yifat Merbl | | 和 | KJ Mu ...
跨界对话:如果向100年后寄一个时间胶囊,你会寄什么?
Xin Lang Cai Jing· 2025-12-08 09:59
Core Insights - The 2025 Influence Entrepreneur Annual Conference, hosted by China Entrepreneur Magazine, focused on the intersection of AI and the future, featuring discussions among industry leaders and experts [3][34]. Group 1: AI and Human Interaction - The ideal interaction with machines is envisioned as "non-perceptual interaction," where machines can anticipate human judgments rather than merely responding to commands [4][35]. - The acceptance and effective use of AI can alleviate identity anxiety, allowing individuals to better understand themselves [4][35]. - There is a call for regulatory measures to manage AI's potential for addiction, as human appreciation systems are vulnerable to manipulation [4][35][30]. Group 2: Future Predictions - By 2030, the world is expected to enter the era of AGI (Artificial General Intelligence), leading to exponential technological advancements and a deep integration of human and machine capabilities [7][38]. - The future may see the emergence of brain-machine interfaces that allow for direct communication of thoughts to machines, fundamentally changing human interaction [9][41]. - The potential for AI to translate animal languages raises intriguing possibilities for interspecies communication by 2040 [9][41]. Group 3: Education and Adaptation - Current educational approaches must evolve to address the integration of AI in creative processes, focusing on how to use AI to enhance rather than replace human creativity [12][43]. - The historical context of technological advancements shows that humans have a strong capacity for adaptation, even in the face of significant changes like the advent of AGI [13][44]. - The need for individuals to manage multiple AI agents may lead to new psychological challenges, highlighting the urgency for educational reform [12][43]. Group 4: Philosophical and Ethical Considerations - The relationship between humans and machines is evolving, with future generations likely to view machines as integral parts of their existence, reducing identity anxiety [11][42]. - Concerns about AI's integration into military applications emphasize the need for safeguards to prevent potential threats to humanity [30][28]. - The discussion includes the importance of enhancing human cognitive abilities through advanced technologies, rather than solely focusing on making machines smarter [52][53]. Group 5: Legacy and Future Vision - Participants expressed a desire to leave behind artifacts for future generations, such as technological milestones that signify the beginning of the AI era [57][58]. - The concept of creating a "time capsule" of current knowledge and data aims to provide future societies with a reference point for rebuilding or understanding their past [58][58].
云知声CEO黄伟斩获2025年度“安永企业家奖”殊荣,诠释新时代企业家精神
Sou Hu Cai Jing· 2025-12-08 07:21
Core Insights - The article highlights the recognition of Huang Wei, CEO of Yunzhisheng, as the recipient of the "Ernst & Young Entrepreneur Award 2025" for his significant contributions to the commercialization of General Artificial Intelligence (AGI) and his exceptional leadership [1][5][10] - The award marks the 20th anniversary of the event and serves as a tribute to the new generation of Chinese tech entrepreneurs who empower industrial upgrades through innovative productivity [1][5] Company Overview - Yunzhisheng, founded in 2012, is one of the pioneers in the research and industrial application of AI voice and semantic technologies in China, leading the way in the commercialization of deep learning technologies across various sectors such as healthcare, home automation, transportation, and government [5][9] - The company has developed the "Shanhai Model," which has demonstrated leading capabilities in global assessments, ranking at the top in both general and specialized medical evaluations, showcasing its technical versatility and domain adaptability [7][9] Industry Context - The article discusses the transition from AI 1.0 to AI 2.0, emphasizing the need for companies to adapt their competitive strategies to focus on rapid learning, foresight, and agile responses rather than solely on technical prowess [9] - Huang Wei's perspective on the AI industry's "bubble" is that it acts as a catalyst for technological transformation, attracting resources and talent to core technologies, ultimately driving industrial upgrades [9][10] Entrepreneurial Spirit - The narrative emphasizes the evolving role of modern entrepreneurs, with Huang Wei stating that the current generation must address the question of strength in addition to existence, highlighting the importance of social value alongside commercial success [7][10] - The article reflects on the broader implications of Huang Wei's achievements as a representation of the entrepreneurial spirit in China's private economy, advocating for foresight, resilience against superficial trends, and a commitment to societal contributions [10]
速递|OpenAI的“红色警报”与秘密武器:ChatGPT产品掌门人尼克·特利,如何引领ChatGPT穿越激流
Z Potentials· 2025-12-08 02:43
尼克 ·特利, Turley 这位曾任 Instacart 产品经理、现掌管 ChatGPT 的高管,正处在前线位置, OpenAI 试图借此保持对谷歌的领先优势。 OpenAI 首席执行官萨姆 ·阿尔特曼近期宣布进入 " 红色警戒 " 状态,旨在激励这家 AI 初创公司的员工对其旗舰产品 ChatGPT 进行更大改进。 这项行动 能否成功,可能取决于鲜为人知的 OpenAI 聊天机器人业务负责人尼克·特利。 30 岁的特利三年前加入 OpenAI ,此前曾在 Dropbox 和 Instacart 任职,此后迅速崛起成为这家 AI 初创公司最具影响力的领导者之一。 他是负责 ChatGPT 及其新产品用户体验的核心管理者之一。在支持者眼中,特利擅长快速推动产品落地,并能协调 OpenAI 内部不同派系的研究人员和工程师达成 共识。 "如果没有尼克, ChatGPT 很可能不会问世,"曾在 ChatGPT 发布期间与特利共事、现任 OpenAI Labs 研究负责人的乔安妮·张表示 然而,对于批评他的许多人 ——其中大多来自 OpenAI 的研究团队来说,特利脱离了与 AI 模型培训及其他对 ChatGPT ...
下一个十年的AI发展图景
Zhong Guo Qing Nian Bao· 2025-12-07 22:52
今年8月,国务院印发《关于深入实施"人工智能+"行动的意见》为我国推动人工智能与经济社会各行业 各领域广泛深度融合提供了指引。10月28日发布的《中共中央关于制定国民经济和社会发展第十五个五 年规划的建议》中再次明确:"深入推进数字中国建设""加快人工智能等数智技术创新""全面实施'人工 智能+'行动,以人工智能引领科研范式变革,加强人工智能同产业发展、文化建设、民生保障、社会治 理相结合,抢占人工智能产业应用制高点,全方位赋能千行百业"。 从线上人工智能(以下简称"AI")大模型与教育、医疗、金融等各行各业深度绑定,持续刷新行业效率 上限;到线下具身智能机器人在工厂协作、社区养老、家庭服务中崭露头角,为人类生产生活带来无限 可能……2025年,人工智能正以前所未有的速度穿透虚拟与现实、串联技术与产业,也让人们不禁畅 想,AI技术的未来发展还能带来怎样的惊喜。 不过,在AI技术打破人机边界的背后,安全治理也成为不可忽视的命题。姚期智提醒,AI算法潜在的 不可靠性可能引发隐私泄露、冲击社会价值伦理等风险。"目前中外正在探索将AI与密码学、博弈学等 理论结合的交叉领域,凝聚国际共识,携手构建AI治理协议。"姚期智 ...