导语：ChatGPT让Meta乱了阵脚。 金融时报ft访谈了杨立昆（ Yann LeCun） ： 2025年初的巴黎大皇宫，马克龙正在主办一场国际AI峰会。 会场外不远处的米其林星级餐厅Pavyllon里。 Yann LeCun坐在两棵塑料圣诞树之间，戴着标志性的黑色雷朋眼镜，看起来精神不错。 这位被称为现代AI"教父"之一的科学家，刚刚结束了在巴黎一周的密集行程。 过去这些天里，他不是在策划什么"世界统治"，用他自己的话说，是在推进"全球辅助"，或者叫"智能放大"。 毕竟，"统治"这个词放在AI语境下，听起来确实有点吓人。 一条新闻打乱的节奏 上次公开露面时，LeCun还是Meta的首席AI科学家，正在展示他对超级智能机器的最新构想。 现在，他准备离开这家工作了十多年的公司，为一家新创公司融资，要把那个构想真正做出来。 英国《金融时报》抢先报道了他的离职消息。 "这基本上逼着我们把时间表往前赶了，" LeCun说。 消息出来后，马克龙给他发了WhatsApp。 虽然LeCun不愿透露总统说了什么，但能看出来，对方挺高兴这家新的"全球性"公司会跟法国保持紧密关系。 LeCun不会去当新公司的CEO，而是做执行 ...

杨立昆不会担任该公司的首席执行官，而是执行董事长，这让他能拥有与在Meta时相同的研究自由。 报道称，杨立昆的新公司名为先进机器智能实验室（Advanced Machine Intelligence Labs），将由法国医 疗人工智能初创公司Nabla的联合创始人兼首席执行官亚历克斯·勒布伦（Alex LeBrun）领导。 自爆出将离开Meta的消息后，杨立昆（Yann LeCun）的日程安排得极其紧凑。"这基本上迫使我们加快 了日程，"他说。报道出炉后，法国总统马克龙给他发了一条WhatsApp消息。杨立昆不愿透露总统具体 说了什么，但暗示总统对这家新的"全球性"公司将与法国保持紧密联系感到高兴。 "我是一名科学家，一个有远见的人。我能激励人们去研究有趣的事物。我很擅长判断哪种技术可行、 哪种不可行。但我做不了首席执行官，"杨立昆说，"我既太缺乏条理，而且也年纪太大了！" 数十年来，他一直在潜心构思如何让机器具备这样的智能。他也毫不讳言自己对大型语言模型的质疑态 度，硅谷当下正痴迷于这类模型，认为它们有望实现超越人类的智能，但在他看来，大型语言模型虽有 实用价值，其本质却受限于语言载体，存在根本性局限。要实 ...

Core Insights - The return of Jeff Bezos and Yann LeCun to the AI sector marks a significant shift in the industry, with their contrasting approaches aiming to address the real bottlenecks in AI technology and its application in creating tangible value [1][3][4] Group 1: Major Players Re-entering the AI Arena - Jeff Bezos has taken on a leadership role in the AI startup "Project Prometheus," securing $6.2 billion in funding, making it one of the best-funded early-stage AI companies globally [4][5] - Yann LeCun, a Turing Award winner, is establishing a new company focused on Advanced Machine Intelligence (AMI), with Meta as a strategic partner, emphasizing foundational research over immediate commercialization [5][6] Group 2: Diverging Paths in AI Development - Bezos's "physical AI" approach targets the optimization of engineering manufacturing in sectors like hardware, automotive, and aerospace, aiming to reduce production cycles significantly [5][7] - LeCun's focus on AMI seeks to address the fundamental challenges of AI, such as understanding the physical world and developing reasoning capabilities, which he believes are essential for the next AI revolution [8][9] Group 3: Capital and Talent Dynamics - The influx of capital into the AI sector is accelerating, with 57% of new unicorns being AI companies, and the funding environment becoming increasingly competitive [10][11] - Talent acquisition has intensified, with companies offering substantial compensation packages to attract top AI researchers, further reshaping the competitive landscape [11][12] Group 4: Industry Trends and Future Outlook - The AI industry is transitioning from a phase of technological explosion to one of deep industry engagement, characterized by a focus on foundational innovation and vertical integration [9][10] - The potential for AI to drive significant advancements in manufacturing and healthcare is evident, with applications already demonstrating substantial efficiency gains and cost reductions [13][14] Group 5: Balancing Opportunities and Risks - While the enthusiasm for AI's potential is high, concerns about a possible bubble due to overvaluation and a lack of sustainable business models are emerging [14][16] - The industry's future will depend on maintaining a balance between innovation quality and commercial viability, as well as navigating regulatory uncertainties [14][16]

Group 1: Generative AI Developments - OpenAI launched two new models, GPT-5.1 Pro and GPT-5.1-Codex-Max, with the former focusing on emotional and intellectual capabilities, while the latter is the first coding model to support a "compression" mechanism [1] - GPT-5.1-Codex-Max can autonomously work for over 24 hours, processing millions of tokens, with a 30% reduction in thinking tokens compared to previous versions, achieving a score of 77.9% on SWE-bench Verified [1] - Internal tests show that 95% of OpenAI engineers use Codex weekly, leading to a 70% increase in team Pull Request numbers [1] Group 2: Image Generation and Processing - Google introduced the Gemini 3 Pro Image preview, a reasoning model that performs internal reasoning before generating images [2] - This model supports 64K input tokens and 32K output tokens, capable of producing images with resolutions between 1K and 4K, and can combine up to 14 input images into one output [2] - It integrates Google search capabilities for up-to-date knowledge, excelling in complex multi-turn image generation editing and high factual accuracy creative tasks [2] Group 3: 3D Technology Advancements - Meta released the SAM 3D family, including SAM 3D Objects and SAM 3D Body, which can convert 2D image segmentation results into 3D models, even in the presence of occlusions [3] - SAM 3D features concept segmentation capabilities, achieving a 47.0% accuracy in the LVIS zero-shot segmentation task, surpassing the previous state-of-the-art (SOTA) of 38.5% [3] - SAM 3D Objects utilizes a 1.2 billion parameter flow-matching Transformer, outperforming other leading models by at least five times in direct comparison tests with human users [3] Group 4: Browser Innovations - QQ Browser's new version v19.8.5 introduces intelligent tab grouping and AI features for multitasking without interference [4] - The new web podcast feature supports AI podcasts and native reading with smart switching, allowing for precise 15-second navigation and five-speed adjustments [4] - The menu and functionality areas have been upgraded, with common tools like bookmarks and history easily accessible at the top, and all fixed function modules supporting drag-and-drop sorting [4] Group 5: Digital Identity Solutions - Second Me provides users with an independent ID and domain in the digital world, acting as an "AI ID" for expression and communication [5] - The product uses AI for precise matching of interests, focusing on finding individuals with detailed similarities rather than just shared interests, reducing communication costs in the industry [5] - Users can record fragmented notes and ideas, allowing their digital persona to continuously add memories and feedback for more natural and accurate expression [5] Group 6: Smart Wearable Technology - Lumia launched the world's first smart earrings, Lumia 2, weighing less than 1 gram and being five times smaller than AirPods, capable of real-time monitoring of head blood flow [7] - The product includes features for tracking sleep, body temperature, menstrual cycles, and overall health status, utilizing patented SwitchBack technology for compatibility with any earrings [7] - Lumia secured an additional $7 million in investment and $5.1 million in government funding, bringing total financing to $17.2 million, with its blood flow tracking technology published in top peer-reviewed journals [7] Group 7: AI Research and Development - Yann LeCun announced his departure from Meta after 12 years to pursue entrepreneurship focused on advanced machine intelligence (AMI) [8] - The new company's goal is to drive the next major AI revolution, enabling systems to understand the physical world, possess long-term memory, and exhibit reasoning capabilities [8] - Meta will partner with the new company, as LeCun emphasizes the importance of world model research, arguing that large language models (LLMs) cannot truly understand the physical world [8] Group 8: Space Computing Initiatives - NVIDIA has sent its H100 GPU into space for the first time, while Google plans to launch 81 satellites equipped with TPUs by 2027, intensifying the space computing competition [9] - China's CAS Tian-Suan has initiated the "Tian-Suan Plan," aiming to deploy a mega-scale space supercomputing center in sun-synchronous orbit, consisting of energy, computing, and communication modules [9] - By mid-2026, CAS Tian-Suan aims to achieve its first GPU supercomputing node in space, targeting a total computing power of 10 EOPS, powered by over 100MW of zero-carbon energy through flexible photovoltaic arrays [9] Group 9: AI Market Insights - NVIDIA reported a record Q3 revenue of $57 billion, with data center business revenue soaring 66% year-over-year to $51.2 billion, and provided a revenue guidance of $65 billion for the next quarter [10] - Jensen Huang refuted the "AI bubble" theory, highlighting a historic shift in computing paradigms from general CPUs to accelerated GPUs, with genuine and sustained demand for computing power [10] - The proportion of GPU-accelerated computing in the global TOP500 supercomputing list surged from 10% six years ago to 90%, with NVIDIA's gross margin around 70%, and global AI infrastructure investment projected to reach $3-4 trillion by 2030 [10]

Group 1 - GAC Toyota's sales vice president denied rumors about discontinuing multiple fuel vehicle models, stating the information is false [1] - The company is reportedly facing market speculation regarding the potential discontinuation of two main fuel vehicle models by 2026 [1] Group 2 - Ant Group's AI assistant "Lingguang" achieved over 500,000 downloads within three days of its launch, ranking seventh on the App Store overall chart and first in the free tools category [2] Group 3 - Airbus announced the initiation of the second phase of its share buyback program, which aims to support future employee stock ownership and incentive plans, with a maximum buyback limit of 4.14 million shares [3] Group 4 - Yann LeCun, Meta's chief scientist, announced his departure to start a new company focused on Advanced Machine Intelligence, aiming to revolutionize AI capabilities [4] Group 5 - Xiaomi's automotive division announced the production of its 500,000th vehicle, with expectations to exceed 400,000 deliveries in 2025 [5] Group 6 - The U.S. government plans to launch an initiative called "Genesis Mission" to enhance AI development, emphasizing its importance comparable to historical projects like the Manhattan Project [6] Group 7 - Google plans to invest $2 billion in building a data center in Turkey, as announced by the Turkish Vice President [7] Group 8 - Singapore-based fintech company RockFlow completed a new funding round of several million dollars, led by Ant Group [9] Group 9 - The Thai Ministry of Finance is considering increasing the VAT rate from 7% to 8.5% in 2028, contingent on economic recovery [10] Group 10 - The Chinese Chemical and Physical Power Industry Association is set to issue a notice to regulate pricing in the lithium iron phosphate industry, aiming to combat ongoing losses and price competition [11] Group 11 - The Ministry of Finance and the Ministry of Industry and Information Technology in China are seeking public opinions on the draft standards for government procurement of new energy vehicles [12] Group 12 - Indonesia is in discussions with Singapore and Malaysia to establish a trilateral data center collaboration [13] Group 13 - China's first green hydrogen coal chemical project has commenced market operations, showcasing a successful model for the green transformation of the coal chemical industry [14]

Core Insights - Yann LeCun, Meta's Chief AI Scientist, will leave the company to establish his own AI startup focused on world models, a field he has extensively researched [2][3] - Meta plans to collaborate with LeCun's startup, aiming to leverage its innovative outcomes [3][4] - LeCun's departure is significant for Meta, as he is regarded as a foundational figure in modern AI, having co-founded the Facebook AI Research (FAIR) and received the Turing Award [5] Group 1: Company Developments - Meta's current AI focus has shifted towards large language models (LLMs), including the Llama series, following a series of setbacks earlier this year, such as the delayed release of the Llama 4 model [4][5] - The company has invested billions in recruiting talent and establishing the Meta Superintelligence Lab (MSL), led by notable figures from Scale AI and GitHub [4] Group 2: Research Focus - LeCun's new startup aims to advance research in advanced machine intelligence (AMI), which he believes will have profound impacts across various economic sectors, some of which overlap with Meta's interests [5] - The startup will pursue the development of systems capable of understanding the physical world, possessing persistent memory, reasoning, and planning complex behavior sequences [3][5]

Core Viewpoint - Yann LeCun, a Turing Award winner, has announced his departure from Meta to start a new company focused on Advanced Machine Intelligence (AMI), aiming to revolutionize AI by enabling systems to understand the physical world, possess long-term memory, reason, and plan complex actions [1][8][14]. Group 1: Company Transition - LeCun's new venture will continue his research on "world models," which he believes are essential for AI to truly understand the physical world [8][27]. - Meta will act as a partner to LeCun's new company, supporting the AMI initiative, which has overlapping interests with Meta's business but also extends into other areas [8][28]. - The departure marks a significant shift in the AI landscape, as LeCun leaves a position he helped establish at Meta's FAIR (Facebook AI Research) amid internal cultural conflicts and strategic misalignments [17][27]. Group 2: Research Focus - The goal of the new company is to drive a major revolution in AI, focusing on systems that can understand the physical world and plan actions without extensive trial and error [8][24]. - LeCun has been a critic of large language models (LLMs), arguing that they lack true understanding of the physical world, and he aims to develop AI that can reason and plan using world models [19][27]. - Recent research contributions include the JEPA theory, which aims to create organized and actionable high-dimensional embedding spaces, seen as a potential pathway to achieving world models [25][27]. Group 3: Industry Impact - LeCun's transition to entrepreneurship at the age of 65 signifies a new exploration phase in AI, moving away from the constraints of corporate environments to pursue foundational scientific challenges [14][27]. - The departure of LeCun, alongside other key figures like Soumith Chintala, indicates the end of an era for Meta AI, highlighting the ongoing evolution within the AI research community [28].

Core Insights - Yann LeCun, a prominent figure in AI and Meta's Chief AI Scientist, announced his departure from Meta to establish a new AI company focused on Advanced Machine Intelligence (AMI) [1] - Meta plans to collaborate with LeCun's new venture, allowing the company to utilize the innovations developed by this startup [1] - The new company aims to create systems that can understand the physical world, possess long-term memory, reasoning capabilities, and plan complex sequences of actions [1] Company Developments - LeCun's new startup will continue the research he has been conducting at Meta and New York University [1] - The technology developed by this new entity is expected to have broad applications across various economic sectors, some of which may align with Meta's business interests, while others may not [1] - The establishment of an independent entity is seen as a way to maximize the widespread impact of Advanced Machine Intelligence [1]

Core Viewpoint - Meta has launched its new "world model" V-JEPA 2, aimed at enhancing AI's physical reasoning capabilities for better understanding and predicting the physical world [1][3][11] Group 1: V-JEPA 2 Overview - V-JEPA 2 is described as a "realistic abstract digital twin" that enables AI to predict the consequences of its actions and plan accordingly [1][3] - The model is 30 times faster than Nvidia's Cosmos model and has been open-sourced for developers to access and integrate into various applications [1][6][5] - V-JEPA 2 builds on the previous V-JEPA model released by Meta, further improving understanding and prediction capabilities [4] Group 2: AI Capabilities - The model provides AI with three core abilities: understanding, predicting, and planning, allowing it to create realistic internal simulations [3][17] - V-JEPA 2 can perform reasoning without the need for labeled video segments, distinguishing it from existing generative AI systems like ChatGPT [3][4] Group 3: Applications and Impact - The model is designed for real-time spatial understanding in AI-driven technologies such as autonomous vehicles, warehouse robots, and drone delivery systems [3][5] - Meta anticipates that V-JEPA 2 will pave the way for AI to operate autonomously in unfamiliar environments, potentially impacting sectors like healthcare, agriculture, and disaster response [18][19] Group 4: Competitive Landscape - The release of V-JEPA 2 is seen as a critical milestone in Meta's long-term AI roadmap, especially in the context of increasing competition with OpenAI, Microsoft, and Google [11][13] - The growing importance of world models in AI research is highlighted, with other companies like Google DeepMind also exploring similar projects [19] Group 5: Leadership and Strategy - Yann LeCun, Meta's Chief AI Scientist, emphasizes the need for AI to build models of how the world operates rather than merely mimicking human text [8][9] - Meta's CEO Mark Zuckerberg is reportedly taking a more hands-on approach to AI development, including significant investments in AI training data and the formation of new teams focused on achieving "superintelligence" [13][14][15]

责编 | 王启隆 出品丨AI 科技大本营（ID：rgznai100） 黄教主的演讲 感觉才没过几天，今年的 GTC 英伟达大会也即将迎来尾声了。 而今年比尔·达利则是对话"AI 教父" 杨立昆 （Yann LeCun），很有前后呼应的感觉。 但 GTC 并不只有黄仁勋和杨立昆，还有许多精彩的演讲与对话，比方说： ………… 接下来的一段时间， CSDN AI 科技大本营 将会在「 GTC 2025 大师谈 」栏目持续更新这些精华内容的全文整理，尽情期待。 比尔·达利 自己就在采访杨立昆之后进行了一场 演讲 ，系统性地讲解了英伟达 2024 一整年的四大项目进展，内容干货很多； OpenAI o1 作者 诺姆·布朗 （Noam Brown）和英伟达的 AI 科学家来了一场 对话 ，他认为现在 AI 圈最需要来一场革命的，就是这些五花八 门的 基准测试 （Benchmark），而且改这个东西还不需要花太多算力资源； 2018 年诺贝尔化学奖得主 弗朗西斯·阿诺德 （Frances Arnold）围绕 AI for Sciense 还有蛋白质工程进行了一场相当硬核的 圆桌对话 ； UC 伯克利教授 彼得·阿比尔 （P ...