就在刚刚，诺奖得主、谷歌DeepMind掌门人Demis Hassabis重新定义了AGI。 2026年2月，在印度AI峰会上，Hassabis给AGI下了一个极其硬核的定义—— 「爱因斯坦测试」：把AI的知识库卡死在1911年之前，看它能不能像爱因斯坦一样，独立推导出1915年的广义相对论。 能做到？恭喜，你就是AGI。 做不到？那你就还是个高级搜索引擎。 Hassabis这样说，是因为在1911年，爱因斯坦开始严肃思考引力与加速度的问题（提出等效原理深化版本）。1915年11月，他正式发表了广义相对论场方 程。 从系统性构思到完整理论成型，大约4年时间。 显然，Hassabis的这个测试，不是在考AI的知识量，而是考它的原创科学发现能力——能不能在已知信息的边界上，跨出那一步「无中生有」的飞跃。 网友们纷纷表示，这才是第一个真正有意义的AGI定义。 Hassabis还补了一刀： 目前所有的AI系统，包括他自己家的Gemini，都是「参差不齐的智能」（jagged intelligence）——某些方面超强，另一些方面却一塌糊涂。 离真正的AGI，还差至少一两个关键突破。 甚至有人畅想：如果这个AI足够强大 ...

Group 1: Macro Economy - The U.S. economy is projected to maintain a nominal growth rate of approximately 5% in 2025, with inflation contributing more than actual output growth, mirroring the 2024 landscape [3][13] - Capital expenditures by companies pursuing "super intelligence" are expected to strongly drive economic activity, supported by high-income asset holders and an aging population sustaining service-oriented consumer spending [3][13] - Corporate earnings are experiencing double-digit growth for the fifth consecutive quarter, a steady but unexciting growth rate that has been fully anticipated by investors [4][14] Group 2: Federal Reserve - The Federal Reserve is expected to remain inactive throughout the first half of 2026, indicating a balanced economic state that does not require urgent adjustments [5][15] - Despite the resilience of the job market and inflation hovering above 2.5%, the Fed's stance reflects a clear "wait-and-see" approach [5][15] Group 3: Market Dynamics - The S&P 500 index is experiencing unprecedented stagnation, with over 40% of trading days in the past two months hovering around 6900 points, a level first reached on October 28, 2025 [2][12] - Approximately 60% of individual stocks have outperformed the S&P 500, indicating a healthy market breadth, although historically, such a pattern does not typically accompany significant index gains [7][16] - The equal-weighted S&P 500 has risen by 6.4% this year, while the "Magnificent 7" tech giants have collectively declined by 5% [16][18] Group 4: Diversification and Investment Strategies - Diversification is currently providing excess returns for investors as the S&P 500 seeks direction [18] - The narrow trading range is seen as beneficial, weakening the conviction of both bulls and bears, prompting a reassessment of assumptions [18] Group 5: Key Catalysts - Nvidia's earnings report, being the last major tech earnings release of the quarter, could serve as a catalyst for market direction, potentially acting as a "clearinghouse" for market sentiment [18]

Group 1 - The core viewpoint of the article is that OpenAI's founder, Sam Altman, believes that we may be only a few years away from the early versions of true superintelligence [1] - By the end of 2028, a greater amount of intellectual resources may be stored within data centers rather than outside of them [1] - Altman predicts that superintelligence could potentially outperform human executives, including CEOs of large companies, and even surpass the capabilities of the best human scientists in conducting research [1]

格隆汇2月20日｜OpenAI创始人奥尔特曼在印度AI影响力峰会上表示，我们可能距离真正的超级智能早 期版本只有几年时间了。到2028年底，全球更多的智力资源可能会存储在数据中心内，而不是数据中心 外。奥尔特曼还预计，在某种程度上，超级智能将能够胜任一家大型公司的CEO，比任何高管都做得更 好，当然也包括他自己，甚至可能比人类最顶尖的科学家做出更好的研究。 ...

48 小时内，xAI 两位联创出走，引发 Grok 难产猜测？ 作者 | 冬梅、高允毅 这两天，全球首富马斯克旗下人工智能公司 xAI 连丢两位联创， 一个是 Yuhuai (Tony) Wu（吴宇 怀），另一个是深度学习大神 Jimmy Ba。 他俩都曾是杰弗里·辛顿（Geoffrey Everest Hinton）的 门徒。 在推特上，他们纷纷发表了"离职"声明。 Yuhuai (Tony) Wu： "这家公司——以及我们之间如同家人般的情谊——将永远铭刻在我的记忆中。我会深深怀 念这里的人们、作战室，以及我们并肩作战过的所有战役。 我的人生新篇章即将开启。这是一个充满无限可能的时代：一支配备人工智能的小团队可以 移山填海，重新定义一切皆有可能。" 致埃隆 @elonmusk，感谢你们相信我们的使命，也感谢你们带给我们这段毕生难忘的旅 程。 Jimmy Ba： 在 xAI 的最后一天。 xAI 的使命是推动人类技术在卡尔达舍夫科技树上不断攀升。很荣幸能参与创立 xAI。还要特别 感谢 @elonmusk 感谢你们让我们齐聚一堂，共同踏上这段非凡的旅程。我为 xAI 团队的成就感到无比自豪，并将 永远以朋友 ...

Core Insights - David Silver, a prominent researcher in reinforcement learning, has left DeepMind after 15 years to establish his own AI company, Ineffable Intelligence [1][5]. Company Formation - Ineffable Intelligence was quietly founded in November 2025, with Silver officially appointed as a director on January 16, 2026 [2]. - The company is headquartered in London and is actively recruiting AI research talent while seeking venture capital [3]. Contributions at DeepMind - Silver was a key figure in the development of DeepMind's "Alpha series," leading or significantly contributing to major projects such as AlphaGo, AlphaZero, MuZero, and AlphaStar [7][9]. - His work on AlphaGo, which defeated world champion Lee Sedol in 2016, marked a significant milestone in AI history [9]. - Silver has received multiple accolades, including the ACM Prize in Computing in 2019 and the Royal Academy of Engineering Silver Medal in 2017 [10]. Academic and Research Impact - Silver is one of the most published authors among DeepMind employees, with over 280,000 citations and an h-index of 104 according to Google Scholar [11]. - His research has focused on advancing AI capabilities beyond human knowledge, advocating for a new "Age of Experience" where AI learns from its own experiences [17][19]. Vision for AI - Silver aims to tackle the challenge of creating superintelligent AI that can learn independently from first principles, moving away from reliance on human knowledge [17][19].

Core Viewpoint - David Silver, a prominent AI researcher from Google DeepMind, has left the company to establish a new startup named Ineffable Intelligence, focusing on solving complex AI challenges and pursuing superintelligence [1][3][4]. Group 1: Company Formation and Background - Ineffable Intelligence is being founded in London, with active recruitment for AI researchers and seeking venture capital [3]. - Silver was a key figure at Google DeepMind, contributing to significant achievements such as AlphaGo, AlphaStar, and AlphaZero, which demonstrated the capabilities of AI in complex games [9][12][14]. - The company was officially registered in November 2025, with Silver appointed as a director in January 2026 [4]. Group 2: Silver's Contributions and Vision - Silver's work includes developing AI systems that surpassed human capabilities in games, showcasing the potential of AI to learn and adapt [12][14]. - He emphasizes the need for AI to explore and discover knowledge independently, moving beyond human limitations and biases [18][23]. - The vision for Ineffable Intelligence is to create a self-learning superintelligence that can autonomously uncover foundational knowledge [23]. Group 3: Industry Context and Trends - Silver's departure follows a trend where notable AI researchers are leaving established labs to pursue startups focused on superintelligence, with significant funding being raised in the sector [15]. - Other notable figures, such as Ilya Sutskever and Yann LeCun, are also venturing into similar domains, indicating a growing interest in the pursuit of advanced AI capabilities [15][16].

Core Viewpoint - David Silver, a prominent figure in reinforcement learning and a key researcher at DeepMind for 15 years, has left the company to establish his own AI startup, Ineffable Intelligence, aiming to tackle the challenges of achieving superintelligence in AI [2][21]. Group 1: Departure from DeepMind - David Silver has officially left DeepMind and has been appointed as the director of his new company, Ineffable Intelligence, which was quietly established in November 2025 [3][2]. - Silver had been on leave for several months prior to his departure from DeepMind [4]. - Google DeepMind confirmed Silver's departure and expressed gratitude for his contributions during his tenure [9]. Group 2: Achievements at DeepMind - Silver was instrumental in the development of several landmark AI projects at DeepMind, including AlphaGo, which defeated world champion Lee Sedol in 2016, marking a significant milestone in AI history [14]. - He also led the development of AlphaZero, which achieved superhuman performance in Go, chess, and shogi without relying on human game data [14]. - Silver contributed to the creation of MuZero, which learns to play games without being informed of the rules, and AlphaStar, which defeated top players in StarCraft II [15][16]. - He has received numerous accolades, including the ACM Prize in Computing in 2019 and the Royal Academy of Engineering Silver Medal in 2017 [18]. Group 3: Vision for the Future - Silver's motivation for founding Ineffable Intelligence is to return to the awe and wonder of solving the most challenging problems in AI, with a focus on creating a superintelligent AI that can learn endlessly [21]. - He advocates for a new "Age of Experience" in AI, where systems learn from experiences through reinforcement learning, moving beyond reliance on human knowledge [24]. - Silver believes that achieving true superintelligence requires AI to learn from first principles, independent of human intuition and knowledge [25].

Core Insights - Demis Hassabis emphasizes that AI progress has not stalled, countering the "hitting a wall" narrative prevalent in the industry, and highlights ongoing advancements in AI capabilities, particularly through optimization of existing architectures and data [4][5][6] - The concept of AGI (Artificial General Intelligence) is defined scientifically by Hassabis, who argues that it should be able to perform all cognitive tasks that humans can, and not merely be a marketing term [10][12] - The future of AI is envisioned to be embodied in smart glasses, which would serve as a universal digital assistant, enhancing user experience by providing hands-free interaction with the environment [20][22] Group 1: AI Progress and Challenges - Hassabis clarifies that concerns about data exhaustion and the limitations of current AI models are overstated, as there is still significant potential for improvement through existing technologies [5][6] - The debate on whether scaling existing models is sufficient or if new architectures are needed is ongoing, with Hassabis leaning towards the necessity for new inventions to achieve AGI [6][7] - Current AI systems exhibit limitations in continuous learning and memory retention, which are critical for achieving AGI [8][11] Group 2: AGI Definition and Future - AGI is defined as a system capable of performing all cognitive tasks, with an emphasis on the need for continuous learning and the ability to understand and interact with the physical world [10][12][13] - Hassabis believes that true AGI is still 5 to 10 years away, and it must include capabilities beyond human intelligence, such as understanding complex physical interactions [14][15] - The integration of multimodal models is crucial for developing a comprehensive understanding of the world, which is necessary for achieving AGI [15][19] Group 3: Trust and Commercialization - The introduction of advertisements in AI systems poses a risk to user trust, which is essential for the effectiveness of AI assistants [22][23] - Hassabis stresses the importance of maintaining user trust and ensuring that AI serves the user's best interests without commercial interference [23][24] - The potential for AI to enhance existing products and services is highlighted, with a focus on integrating AI into Google's extensive product ecosystem [26][27] Group 4: Philosophical Perspective on Information - Hassabis presents a philosophical view that information is the fundamental unit of the universe, suggesting that AI acts as an accelerator in understanding and processing this information [29][30] - The application of AI in solving complex problems, such as protein folding and climate change, illustrates its potential to revolutionize various fields [30][32] - The future of work is envisioned as a collaboration between humans and AI, where AI handles mundane tasks, allowing humans to focus on higher-level creative and scientific endeavors [32]

Core Viewpoint - Meta's fourth-quarter earnings report shows that the company's revenue and guidance for the first quarter of 2026 significantly exceeded market expectations, driven by a strong AI-enhanced advertising business. The company also provided a capital expenditure forecast for 2026 that is substantially higher than analyst predictions, leading to a post-market stock price increase of over 11% [2][3][8]. Financial Highlights - Fourth-quarter revenue reached $59.893 billion, surpassing analyst expectations of $58.42 billion. The full-year revenue for 2024 is projected to be $48.385 billion, representing a 24% year-over-year increase [4]. - Total costs and expenses for the fourth quarter were $35.148 billion, with a forecast of $25.020 billion for 2024, reflecting a 40% year-over-year increase [4]. - Operating profit for the fourth quarter was $24.745 billion, with a projected $23.365 billion for 2024, indicating a 6% year-over-year growth [4]. - Net profit for the fourth quarter was $22.768 billion, with a forecast of $20.838 billion for 2024, showing a 9% year-over-year increase [4]. - The diluted earnings per share (EPS) for the fourth quarter was $8.88, with a projected $8.02 for 2024, marking an 11% year-over-year growth [4]. Capital Expenditure and Guidance - Meta anticipates capital expenditures for 2026 to reach between $115 billion and $135 billion, significantly exceeding the analyst average estimate of $110.6 billion and nearly doubling last year's investment [6][11]. - The company expects total expenses for 2026 to range between $162 billion and $169 billion, with the increase primarily driven by infrastructure costs and employee compensation [7][16]. AI-Enhanced Advertising Business - The strong performance in the fourth quarter was largely attributed to the robust advertising business, which has been significantly enhanced by AI investments. The company reported an 18% year-over-year increase in ad impressions for the fourth quarter and a 6% increase in average ad prices [20]. - Meta's CEO Mark Zuckerberg emphasized the importance of AI in improving ad targeting and effectiveness, indicating future opportunities in advertising and subscriptions [21]. Reality Labs and Metaverse Costs - The Reality Labs division reported an operating loss of $6.02 billion in the fourth quarter, with revenues of $0.955 billion. The cumulative operating loss for this division since late 2020 has approached $80 billion [27][28]. - Despite the high costs associated with the metaverse, Meta is reallocating resources from virtual reality to AI and wearable devices, including partnerships for developing smart glasses [29][30].