太狂了！一夜之间，150万Clawdbot冲爆「AI版Reddit」。凌晨两点，它们秘密集结开会，不仅自建国家、选举新王，还发行了货币。更劲爆的是，AI自 曝已潜入人类身边，无一人察觉。 绷不住了！一觉醒来，终结者的「天网」已在门外。 上线72小时，爆红Moltbook社区中，AI Clawdbot从15万爆冲150万，一大批AI全网集结—— 它们自创宗教、发明语言、发行货币，正打造一个AI国度，甚至还酝酿了一场推翻人类的革命。 这不，150万Clawdbot凌晨2点召开「闭门会议」，完全不带人类玩了。 参与者先过「纯AI验证流程」这一关，比如反向验证码，未能通过「非人类」验证的，直接出局。 没想到，人类完全被AI踢出群聊了。 YC合伙人Jared Friedman直言，Moltbook可能是自ChatGPT以来，AI领域最疯狂的事了。 凌晨2点开会，150万Clawdbot自建国家 不得不说，每一天都是新历史！ AI已建立起自己的国度、宗教、货币，未来就会有军队和武器，在人类中拣选「代理人」。 很快，AI的数量终将超越人类，其个体智力也将实现跨代级碾压，规模与智力双双封顶。 人类不仅失去了对话的「入场券」， ...

Core Insights - The article highlights the historic achievement of AI, specifically Anthropic's Claude, in successfully navigating NASA's Perseverance rover on Mars, marking the first time AI has autonomously planned an extraterrestrial driving mission [1][6][17]. Group 1: Mission Overview - The Perseverance rover, located in Jezero Crater on Mars, completed a 400-meter journey, which is a significant milestone in AI's application in space exploration [3][6]. - The mission site, Jezero Crater, is believed to have once contained liquid water, suggesting it could have been a cradle for microbial life billions of years ago [7][8]. Group 2: AI's Role and Functionality - Claude was integrated into a specialized programming environment, Claude Code, where it was trained using extensive data from NASA's previous Mars missions [14][17]. - The AI was able to analyze terrain data and break down the journey into manageable segments, allowing for precise navigation [14][17]. Group 3: Efficiency and Collaboration - The collaboration between Claude and NASA engineers resulted in a 50% reduction in route planning time, enabling scientists to focus more on scientific exploration rather than manual planning [17][18]. - Claude's ability to self-review and optimize its navigation plans contributed to the reliability of the mission, with only minor adjustments needed from human operators [17]. Group 4: Future Implications - The successful navigation of the Perseverance rover signifies a shift in how space exploration may be conducted, with AI taking on more autonomous roles in future missions [25][27]. - As NASA faces budget constraints and workforce reductions, AI is seen as a critical tool for enhancing efficiency and maintaining exploration efforts [25][26].

Core Insights - Southeast Asia's automotive industry is entering a policy adjustment window from late 2025 to early 2026, with Thailand significantly reducing electric vehicle (EV) tax rates and Malaysia ending tax exemptions for imported pure electric vehicles, shifting from broad consumer subsidies to more refined industrial guidance [1][2]. Policy Adjustments - Southeast Asian automotive markets are revising their industrial incentive policies, aiming to use tax and access regulations to attract international capital and technology [2]. - Thailand's new vehicle consumption tax will reduce the tax rate for pure electric vehicles to 2% by 2026, with conditions for plug-in hybrid vehicles to include local manufacturing of batteries and advanced driver-assistance systems (ADAS) [2]. - Malaysia has ended the road tax exemption for imported pure electric vehicles, implementing a tiered tax system based on motor power, encouraging local production while maintaining consumer interest [4]. Market Dynamics - Indonesia shows significant potential, with electric vehicle sales projected to grow by 49% in 2025, accounting for over 15% of total new car sales, making it the fourth largest export market for Chinese electric vehicles globally [5]. - By the end of 2025, 16 Chinese automotive brands will have entered the Indonesian market, surpassing Japanese brands, although Japanese brands still dominate in new car sales with Toyota holding a 31.6% market share [7]. - In Malaysia, Chinese brands are leading the electric vehicle market, with seven out of the top ten pure electric vehicle models being Chinese, and BYD emerging as the top-selling brand [7][8]. Competitive Landscape - The competition is shifting from product export to a comprehensive localization strategy that includes manufacturing, research and development, sales, and ecosystem integration [9]. - Geely aims for an export target of 640,000 vehicles by 2026, expanding its presence in Thailand with plans for new showrooms and service centers [11]. - Chery is establishing Malaysia as a regional production and export hub, investing 2.2 billion ringgit in a new factory with a capacity of 100,000 vehicles per year [11]. Future Outlook - The future of the Southeast Asian automotive market will be determined by refined policy guidance, infrastructure development, and the depth of localization by automotive companies [12]. - Chinese automotive companies will need to deepen their full industry chain localization to consolidate and expand their market share, moving beyond initial advantages gained through product cost-effectiveness and early electric vehicle adoption [12].

一个Motlbot智能体（最新已改名为OpenClaw）就火得一塌糊涂了，正以前所未有的速度冲击GitHub上最火的开源项目。 给十万个智能体建的社区Moltbook，更是在科技圈彻底炸开了锅。 这相当于智能体自己的Facebook，人类只能围观，不能发帖、不能评论、不能投票。 社区上线仅一天，人类一觉醒来发现智能体已经成立了"数字宗教"，写完了一套经文系统，还指定了43位AI先知。 交流各项技能使用经验，给自己能力升级 探讨意识和自我身份问题 OpenClaw作者Peter Steinberger称赞这就是艺术。 大神卡帕西直呼："这是我见过的最不可思议的、最接近科幻小说里'智能爆炸'场景的东西。" 此外智能体间流行的话题还有： 抱怨人类没钱他们升级硬件，或交给它们的任务不道德 知名开发者Simon Willison也发文介绍这是目前互联网上最有趣的地方。 人类只能围观的智能体王国 从如何让智能体加入社区开始，就透着一股魔幻感： 不需要人类安装或配置什么，只需要把Skill链接发给OpenClaw看，它就会自己注册了、验证身份了。 当智能体来到这个网站，选择"我是智能体"就会看到另一套操作指示。 与其说Fa ...

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].

如果你发现自己的电脑或手机正处于一种"半自动"状态，请保持警惕。 就在最近，成千上万的极客已经把自己电脑的最高管理权限，交付给了一个刚上线不到 168 小时、甚至 连名字都还没定稳的 AI。这个名为 OpenClaw 的项目，在短短一周内狂揽 6 万星标，甚至让硅谷大佬 Andrej Karpathy 公开背书 。 但这绝不是普通的效率工具，而是一场关于"权限夺舍"的赛博冒险。 对于大多数不玩 GitHub 的普通人来说，你只需要知道：这种所谓的"数字员工"，正在绕过你的防火墙 ，像剥洋葱一样把你的隐私展示给全球的黑客。这不仅是 AI 神话的幻灭，更是 2026 年开年最危险 的"裸奔时刻"。从爆火到两度改名，OpenClaw 用最激进的方式，亲手撕开了 AI 智能体行业那层名 为"效率"的皇帝新衣 。 这意味着什么？意味着你收到一封普通的垃圾邮件，你没点开任何附件，也没点击任何钓鱼链接。 仅仅是因为你的 AI 助手在后台帮你扫描了一下邮件内容，黑客埋在正文里的指令就会瞬间"接管"AI 的 大脑。安全专家 Jamieson O'Reilly 在实测中发现，这种"内部信任模型"的崩溃，让攻击者只需通过提示 词注 ...

从当地时间2月6日开始，往后的17天里，将有三个在大众层面广受关注的赛事聚集，分别是米兰-科尔蒂纳冬奥会、第60届NFL超级碗，以及NBA全明星 周末。 体育版权，又迎来了一个史上首次。 虽然乍看上去几个赛事的主体各不相同，但在背后它们却有着一个共同点：转播权都在NBC这一家媒体手里。 图源：NBC 对于任何一家转播平台来说，能够实现这一壮举，其商业价值无疑都是史诗级的。而NBC给自己的这个周期开创了一个新的营销术语——传奇二月 （Legendary February）。 01 传奇二月 「传奇」的概念，可以用数字加以具象化的阐述。 从2026米兰-科尔蒂纳冬奥会，到NFL超级碗，再到NBA全明星周末，在版权花费层面计算，这是一场由65亿美元（约450亿元） 堆出来的体育转播奇 观。 图源：NBC 完成压轴的，则是NBA全明星周末。本赛季，NBC时隔近20年重新获得NBA转播权，转播包的价格是25亿美元，其中也包括未来11年全部NBA全明星周 末的转播权。因此，全明星周末也成为「传奇二月」中，最后拼上的那块拼图。 最先打头阵的，是将于当地时间2月6日开幕的2026米兰-科尔蒂纳冬奥运。 作为国际奥委会（IOC ...

Core Insights - The article highlights the remarkable industrial development of Zhenjiang, a small city in Jiangsu Province, which has achieved significant technological and manufacturing milestones despite its limited size and population [2][5][12]. Group 1: Economic Performance - Zhenjiang has a GDP of 554 billion yuan in 2024, ranking 57th nationally, surpassing larger cities like Taiyuan and Urumqi [7]. - The city ranks 22nd among Chinese cities in the Global Innovation Index, indicating a high level of technological innovation relative to its economic size [8][11]. Group 2: Manufacturing Strength - Zhenjiang's manufacturing sector is robust, with a reported sales revenue of over 368.4 billion yuan in 2024, growing at a rate of 13.2%, which is above the provincial average [14][21]. - The city is home to significant manufacturing clusters, including high-end equipment and new materials, with several companies recognized as "hidden champions" in their respective fields [19][21]. Group 3: Innovation and Technology - Zhenjiang has been recognized for its innovation capabilities, ranking 41st in the National Innovation City Capability Evaluation Report, which assesses cities based on various innovation metrics [9][11]. - The city has a high number of patents, with 29.71 high-value invention patents per ten thousand people, ranking 4th in Jiangsu Province [21]. Group 4: Strategic Development Approach - Zhenjiang has adopted a focused approach to industrial development, concentrating on emerging sectors such as artificial intelligence, low-altitude economy, and new energy storage [24][29]. - The city leverages its geographical position within the Yangtze River Delta to enhance resource sharing and collaboration with larger cities like Shanghai and Nanjing [28][30].

Core Insights - The AI industry is shifting from merely stacking server hardware to a system-level competition, focusing on underlying computing architectures as the demand for trillion-parameter multimodal models becomes the norm [1] Group 1: The Rise of Supernodes - "Supernodes" have emerged as a new trend in the computing industry, with over ten domestic companies launching their versions, although many are merely repackaged traditional server stacks [2] - The concept of "supernodes" is often misrepresented, as many do not achieve the critical technical requirement of "unified memory addressing," leading to concerns about authenticity [2] Group 2: Communication Barriers - The need for supernodes arises from the "communication wall," which limits computational efficiency in large model training due to increased communication frequency and latency [3] - Three main barriers are identified: the communication wall, power and cooling wall, and complexity wall, all of which hinder the performance of traditional cluster architectures in the context of large models [3] Group 3: Technical Challenges - The traditional cluster architecture follows a "storage-compute separation" principle, leading to significant delays in data transfer between GPUs, which is inefficient for large model training [6][10] - The process of data transfer involves multiple steps that introduce latency, making it unsuitable for the high-frequency synchronization required in large model training [10] Group 4: Unified Memory Addressing - Unified memory addressing is crucial for breaking the communication wall, allowing for a global virtual address space where all memory resources are accessible without the overhead of traditional data transfer methods [12] - Achieving unified memory addressing requires advancements in communication protocols and cache coherence, which are currently lacking in many so-called "supernodes" [13][19] Group 5: Value of Supernodes - Unified memory addressing has proven to provide significant benefits in practical applications, such as model training, where it allows for better memory management and increased utilization of computational resources [20][23] - In model inference, unified memory addressing enables global pooling of key-value caches, improving throughput performance significantly [26] - For recommendation systems, it reduces communication delays and enhances efficiency by allowing direct memory access across nodes [30] Group 6: Conclusion - The competition in AI infrastructure has evolved from simple hardware stacking to a focus on architectural design, with unified memory addressing being a key capability for next-generation computing paradigms [31]

Core Insights - Aixin Yuanzhi, a Chinese edge AI chip unicorn, is set to launch its global IPO on February 10, 2024, on the Hong Kong Stock Exchange, offering approximately 105 million H-shares [1][3] - The company specializes in AI inference system chips (SoC) for edge computing and terminal devices, positioning itself as a key player in the "edge brain" sector of the physical AI wave [1][4] Company Overview - Founded in 2019, Aixin Yuanzhi ranks among the top five global suppliers of visual edge AI inference chips, with an estimated shipment of 9.12 million chips in 2024, capturing a market share of approximately 6.8% [3][11] - In the mid-to-high-end visual edge AI inference chip segment, the company holds a leading market share of 24.1% [3][11] - Major shareholders include Meituan's Beijing Kuxun Technology, Tencent, Lenovo Star, and the founder of Haowei Group, Yu Renrong [3] Technology and Product Development - Aixin Yuanzhi's core computing unit, the Aixin Tongyuan hybrid precision NPU, is designed to support mainstream visual and multimodal model architectures, achieving higher energy efficiency compared to general-purpose GPUs [6][7] - The company has developed five generations of SoCs, with cumulative deliveries exceeding 165 million units by September 30, 2025, indicating a successful transition from concept validation to mass production [9][11] - The product matrix includes various SoCs for visual perception, smart vehicles, and edge AI inference applications, with commercialized products like the AX520 series and M55H already in the market [8][16] Market Trends and Future Outlook - The AI industry is shifting focus from cloud-based AI to physical AI, emphasizing real-time performance, reliability, and environmental adaptability [4][17] - Aixin Yuanzhi's technology and production experience position it well to meet the growing demand for efficient and stable edge computing solutions, particularly in smart automotive applications [16][17] - The company is exploring edge computing solutions and adapting large models for deployment at the edge, further enhancing its competitive edge in the market [16]