Workflow
量子位
icon
Search documents
黄仁勋回应AMD“送股”OpenAI:很高明的交易,OpenAI现在还没钱给我付账
量子位· 2025-10-09 04:52
Core Insights - Huang Renxun expressed surprise at AMD's strategy of exchanging 10% of its shares for OpenAI orders, calling it clever [1][3] - He emphasized that NVIDIA's relationship with OpenAI is fundamentally different, as NVIDIA sells products directly to OpenAI rather than through equity exchange [4] - OpenAI currently lacks the funds to pay for its large orders, needing to raise capital through future revenue growth, equity, or debt financing [5][7] NVIDIA and OpenAI Transactions - NVIDIA has the opportunity to co-invest in OpenAI's future financing rounds, with Huang expressing regret for not investing more when OpenAI was founded [8] - NVIDIA announced plans to invest up to $100 billion in OpenAI, which will build systems requiring 10 gigawatts of power, equivalent to 4 to 5 million GPUs [12][13] - OpenAI has also agreed to purchase AMD chips, committing to a significant procurement of AMD's upcoming MI450 series [14] Circular Trading Dynamics - The transactions create a closed-loop funding mechanism where NVIDIA's investment flows to Oracle through OpenAI, which then purchases NVIDIA hardware [16] - The total value of AI computing transactions between OpenAI, NVIDIA, AMD, and Oracle has surpassed $1 trillion, while OpenAI is projected to struggle with positive cash flow until 2029 [16] Expansion of NVIDIA's AI Investments - NVIDIA confirmed its participation in a $20 billion funding round for Musk's xAI, with plans to invest $2 billion [17] - The investment will utilize a special purpose vehicle (SPV) structure, with funds allocated for purchasing NVIDIA processors [18] - NVIDIA has also invested in CoreWeave, holding a 7% stake, and is actively involved in numerous AI venture capital transactions [19][20] Market Concerns - The intricate web of transactions has raised concerns about a potential AI bubble, with analysts warning that these deals could signal early warning signs if a bubble were to burst [20] - NVIDIA representatives clarified that the company does not require its invested companies to use NVIDIA technology [21][22]
2025人工智能年度评选启动!3大维度5类奖项,正在寻找AI+时代领航者
量子位· 2025-10-09 04:52
组委会 发自 凹非寺 量子位|公众号 QbitAI 为了让更多从业者感受智能浪潮的跃迁,也为了给予更多同行同路人掌声与鼓舞,我们将正式启动 「2025人工智能年度榜单」评选报名 。 这是量子位人工智能年度榜单的 第8年 。八年来,我们见证了技术的突破与落地,产业的融合与重塑,也见证了一批又一批推动时代前行的 企业、人物与产品。 在人工智能重新定义一切的时代里,智能技术已不再是单一工具,而是产业与社会协同进化的驱动力。我们期待通过这场年度评选,去发现并 致敬那些真正引领变革、开拓边界的探索者与实践者。 本次评选将从 企业 、 产品 、 人物 三大维度,设立五类奖项。欢迎企业踊跃报名! 让我们共同见证年度之星,点亮未来的方向。 企业榜 产品榜 人物榜 2025 人工智能年度 焦点人物 详细评选标准及报名方式如下。 2025 人工智能年度领航企业 2025 人工智能年度 领航企业 2025 人工智能年度 潜力创业公司 2025 人工智能年度 杰出产品 2025 人工智能年度 杰出解决方案 评选标准 : 1、注册地在中国,或主营业务主要面向中国市场; 2、主营业务属于人工智能及相关产业,或已将人工智能广泛应用于主营业 ...
备受Meta折磨,LeCun依旧猛发论文!新作:JEPAs不只学特征,还能精准感知数据密度
量子位· 2025-10-09 04:52
Core Insights - The article discusses a new research paper by Yann LeCun's team that reveals the hidden capability of the self-supervised model JEPAs (Joint Embedding Predictive Architecture) to learn data "density" [2][5][6] - This finding challenges the long-held belief that JEPAs only excel at feature extraction and are unrelated to data density [7] Group 1: Key Findings - JEPAs can autonomously learn the commonality of data samples during training, allowing them to assess the typicality of a sample without additional modifications [6][11] - The core discovery is that the anti-collapse mechanism enables precise learning of data density, which was previously underestimated [11][12] - The research highlights that when JEPAs output Gaussian embeddings, they must perceive data density through the Jacobian matrix, making the learning of data density an inherent result of the training process [11] Group 2: Practical Applications - The team introduced a key tool called JEPA-SCORE, which quantifies data density and scores the commonality of samples [14][15] - JEPA-SCORE is versatile and can be applied across various datasets and JEPAs architectures without requiring additional training [16][17] - Experiments demonstrated that JEPA-SCORE effectively identifies typical and rare samples across different datasets, confirming its reliability and general applicability [18] Group 3: Research Team - The research was a collaborative effort involving four core researchers from Meta's FAIR, including Randall Balestriero, Nicolas Ballas, and Michael Rabbat, each with significant backgrounds in AI and deep learning [26][28][30][32][34][36]
更高智商更快思考!蚂蚁开源最新万亿语言模型,多项复杂推理SOTA
量子位· 2025-10-09 04:52
Core Insights - Ant Group has officially released its flagship model, Ling-1T, which boasts one trillion parameters, surpassing both open-source models like DeepSeek-V3.1-Terminus and closed-source models such as GPT-5-main [1][56] - Ling-1T demonstrates state-of-the-art (SOTA) performance in various complex reasoning benchmarks, including code generation and mathematical reasoning [1][3] - The model exhibits impressive reasoning speed, initiating thought processes almost instantaneously upon input [4][60] Performance and Capabilities - Ling-1T achieved optimal performance on the AIME 25 competition mathematics leaderboard, outperforming numerous models [3] - The model can efficiently handle complex logical deductions and generate lengthy texts with smooth output [4][60] - In practical tests, Ling-1T effectively solved a spatial geometry optimization problem by proposing four distinct solutions, each with detailed steps and applicable scenarios [8][9] Technical Innovations - The model's architecture is based on Ling 2.0, with a total parameter count expanded to one trillion, allowing for enhanced information storage and expression [38][41] - The training process involved over 20 trillion tokens of high-quality, reasoning-focused data, supporting a maximum context window of 128K tokens [39][40] - A novel "mid-training + post-training" approach was employed, enhancing the model's reasoning capabilities and efficiency [40][59] Training Methodology - The training was divided into three phases: initial knowledge acquisition, reasoning skill development, and mid-training to prepare for post-training [45][44] - A new learning rate strategy, WSM (Warmup-Stable and Merge), was introduced to optimize training without traditional decay, resulting in improved performance across tasks [49][48] - The LPO (Linguistics-Unit Policy Optimization) method was innovatively applied, allowing for more precise training by using sentences as the optimization unit [52][54] Market Context - The release of Ling-1T positions Ant Group among the leading players in the trillion-parameter open-source model space, alongside Qwen and Kimi [61] - The ongoing trend of rapid advancements in China's open-source model landscape is highlighted, with multiple significant releases from various companies [62][56] - The competitive landscape suggests that further innovations and surprises in the large model sector are likely to emerge from China [63]
首个全自动AI科学家诞生!西湖大学最新成果:性能超越人类SOTA基线183.7%
量子位· 2025-10-08 13:06
△ 对比DeepScientist与人类专家的研究进展 在AI文本检测任务中,DeepScientist仅用两周时间就实施和验证了超过 1000种 不同的假设,在此期间取得了相当于人类三年的进展。 在RAID数据集测试中,DeepScientist设计的方法实现了 7.9% 的AUROC提升,成功 超越了人类现有SOTA方案 。 另外DeepScientist还在智能体失败归因、LLM推理加速等任务上也分别达成了新的SOTA。 DeepScientist团队 投稿 量子位 | 公众号 QbitAI 人类科学家三年的工作量,如今AI两周就能轻松搞定! 最近,来自西湖大学的自然语言处理实验室发布了 DeepScientist 系统,这也是 首个 具有完整科研能力,且在无人工干预下,展现出目标 导向、持续迭代、渐进式超越人类研究者最先进研究成果的AI科学家系统。 下面是更多详细内容介绍。 从"科研助理"到"首席科学家":AI科研模式的变革 过去的AI Scientist系统,如果不给定一个清晰明了的科研目标,就很容易陷入对现有知识的机械组合与无效试探的窠臼中,最终形成的科研 产出在人类专家看来缺乏焦点,科学价值不高 ...
直播预告:光轮智能 × NVIDIA带来Sim2Real关键突破
量子位· 2025-10-08 13:06
Core Viewpoint - The collaboration between Guanglun Intelligent and NVIDIA aims to leverage SimReady and AI to achieve seamless migration from virtual simulation to the physical world, addressing key challenges in robot development and implementation [2][3]. Group 1: Live Broadcast Highlights - The live broadcast will focus on the technological breakthrough of Sim2Real, detailing how both companies utilize SimReady and AI to overcome challenges in robot development [2]. - Experts will share insights on the technological trends and commercialization paths in the fields of robotics and AI, drawing from their practical experiences [4]. Group 2: Collaboration Progress - Exclusive updates on the latest achievements and plans in technology research and application scenarios from the partnership between Guanglun Intelligent and NVIDIA will be disclosed [3]. Group 3: Key Speakers and Event Details - The live broadcast will feature Steve Xie, the founder and CEO of Guanglun Intelligent, and Madison Huang, Senior Director of Product Marketing at NVIDIA [6]. - The event is scheduled for October 9 at 00:00 Beijing time, which corresponds to October 8 at 09:00 Pacific time [6].
30家Tokens吞金兽,每家烧光万亿Tokens!OpenAI最大客户名单曝光,多邻国上榜
量子位· 2025-10-08 04:25
Core Insights - OpenAI has identified 30 companies that have consumed over a trillion tokens, showcasing significant engagement with AI applications [1][3][5] Group 1: Companies Overview - Duolingo is a language learning app known for its gamified course design, boasting over 700 million users and 70 million monthly active users, making it a leading client of OpenAI [10][11] - OpenRouter serves as a multi-model aggregation platform, allowing users to access various AI models through a unified API, positioning itself as a potential monopoly in the API market [15][17] - Canva is an online graphic design platform that has integrated AI to simplify design processes, resulting in high token consumption due to its multi-modal content requirements [21][22] - Perplexity is an AI-native search engine that processes multiple web pages simultaneously, leading to high token usage with over 20 million monthly active users [24][25] Group 2: Token Consumption Insights - High token consumption is attributed to three main factors: frequent user interactions, complex task requirements, and platform effects that aggregate demand for AI services [25][27] - The industry is shifting towards a new benchmark of daily token consumption, with 1 billion tokens per day being seen as a new standard for evaluating AI application viability [28][29][31]
另一位Yao Shunyu也跳槽了:与Anthropic价值观有根本分歧
量子位· 2025-10-08 04:25
Core Insights - The article discusses the recent transition of Shunyu Yao, a prominent AI researcher, from Anthropic to Google DeepMind, highlighting his background and motivations for the move [1][4][41]. Group 1: Background and Career Transition - Shunyu Yao, a distinguished alumnus of Tsinghua University, recently joined Google DeepMind as a Senior Research Scientist after leaving Anthropic, where he contributed to the Claude AI model [1][41]. - Yao's departure from Anthropic was influenced by a fundamental disagreement in values, which he stated accounted for 40% of his decision, while the remaining 60% involved internal details he chose not to disclose [21][24]. - His experience at Anthropic was marked by a high workload, which he described as "super busy," preventing him from reflecting on his transition from physics to AI research until after his departure [7][8][18]. Group 2: Insights on AI Research - Yao expressed that the field of AI research, particularly in large models, is currently in a chaotic state, akin to the early days of thermodynamics, where foundational principles are not yet fully understood [14][15][16]. - He noted the rapid evolution of AI, with the Claude model progressing from version 3.7 to 4.5 within a year, emphasizing the fast-paced nature of advancements in the field [27]. - Yao's background in theoretical physics provided him with a unique perspective on AI research, allowing him to appreciate the ability to identify patterns without fully understanding the underlying principles [16][18]. Group 3: Academic Achievements - During his undergraduate studies, Yao made significant contributions to condensed matter physics, publishing groundbreaking work in the prestigious journal Physical Review Letters [30][31]. - His research achievements include the introduction of new physical concepts and theories related to non-Hermitian systems, which have been recognized as substantial contributions to the field [32][33]. - After completing his PhD at Stanford University, Yao's work continued to focus on cutting-edge topics in quantum mechanics, further establishing his reputation as a leading researcher [35].
2025人工智能年度评选启动!3大维度5类奖项,正在寻找AI+时代领航者
量子位· 2025-10-08 04:25
组委会 发自 凹非寺 量子位|公众号 QbitAI 为了让更多从业者感受智能浪潮的跃迁,也为了给予更多同行同路人掌声与鼓舞,我们将正式启动 「2025人工智能年度榜单」评选报名 。 这是量子位人工智能年度榜单的 第8年 。八年来,我们见证了技术的突破与落地,产业的融合与重塑,也见证了一批又一批推动时代前行的 企业、人物与产品。 在人工智能重新定义一切的时代里,智能技术已不再是单一工具,而是产业与社会协同进化的驱动力。我们期待通过这场年度评选,去发现并 致敬那些真正引领变革、开拓边界的探索者与实践者。 产品榜 人物榜 2025 人工智能年度 焦点人物 详细评选标准及报名方式如下。 2025 人工智能年度领航企业 本次评选将从 企业 、 产品 、 人物 三大维度,设立五类奖项。欢迎企业踊跃报名! 让我们共同见证年度之星,点亮未来的方向。 企业榜 2025 人工智能年度 领航企业 2025 人工智能年度 潜力创业公司 2025 人工智能年度 杰出产品 2025 人工智能年度 杰出解决方案 将面向中国人工智能领域,评选出最具综合实力的企业, 参选条件 : 评选标准 : 2025 人工智能年度潜力创业公司 聚焦于中国人 ...
2025诺贝尔物理学奖颁给了谷歌量子计算机打造者
量子位· 2025-10-07 10:55
Core Viewpoint - The Nobel Prize in Physics 2025 was awarded to three scientists in the field of quantum mechanics: John Clarke, Michel H. Devoret, and John M. Martinis, for their discoveries related to macroscopic quantum tunneling effects and energy quantization phenomena in circuits [1]. Group 1: John Clarke - John Clarke's research focuses on superconductivity and superconducting electronics, particularly in low-temperature physics [4]. - He is best known for inventing and improving the superconducting quantum interference device (SQUID), which is a highly sensitive flux-to-voltage converter used in various fields such as condensed matter physics and medical physics [4]. - Clarke was born in 1942 in Cambridge, UK, and has received numerous awards, including the Fritz London Prize for his contributions to low-temperature physics [7][11]. Group 2: Michel H. Devoret - Michel H. Devoret is recognized as one of the founders of "quantum electronics," focusing on the quantum behavior of electronic systems at the mesoscopic scale [16]. - He has made significant contributions to understanding the fundamental mechanisms of quantum non-equilibrium physics in superconducting circuits, laying a solid foundation for quantum technology [18]. - Devoret has received several prestigious awards, including the 2024 Comstock Prize in Physics and the 2022 Micius Quantum Prize [19]. Group 3: John M. Martinis - John M. Martinis's core contribution to the Nobel Prize was his research on the quantum behavior of the phase difference in Josephson junctions, demonstrating that macroscopic circuit systems can exhibit quantum tunneling and energy level discretization [20]. - He played a pivotal role in achieving "quantum supremacy" with a 53-qubit processor, surpassing the computational power of the world's strongest classical supercomputer [24]. - Martinis has held various prestigious positions, including serving as the Chief Scientist for Quantum Hardware at Google's Quantum AI Lab, and has co-founded companies focused on practical quantum computing [26][28].