11月27-28日，被誉为"年度科技与商业风向标"的36氪WISE2025商业之王大会，在北京798艺术区传导空间落地。 今年的WISE不再是一场传统意义上的行业峰会，而是一次以"科技爽文短剧"为载体的沉浸式体验。从AI重塑硬件边界，到具身智能叩响真实世界的大门； 从出海浪潮中的品牌全球化，到传统行业装上"赛博义肢"——我们还原的不仅是趋势，更是在捕捉在无数次商业实践中磨炼出的真知。 我们将在接下来的内容中，逐帧拆解这些"爽剧"背后的真实逻辑，一起看尽2025年商业的"风景独好"。 光联芯科 CEO 陈超 不能。为什么？因为算力行业面临了两大挑战，有带宽的瓶颈、也有能耗的问题。 以下是真知创投合伙人、光联芯科CEO 陈超先生的演讲实录，经36氪编辑： 大家下午好！我是来自光联芯科的陈超。非常高兴有机会跟大家分享,我的演讲主题是"算力·无界 光互连是通往AGI的必由之路"。 在正式开始之前，我想邀请大家来先看一组图片，这三张图由Open AI Sora多模态大模型生成的，算力规模不同，从左到右别用到了300张GPU、1250张 GPU和10000张GPU，从左到右也是图片质量越来越好。那么，我们是否可以做一个基 ...

Core Viewpoint - The debate over whether AI data center investments are overheated is intensifying in Wall Street and Silicon Valley, with significant capital expenditure plans announced by major tech companies, raising concerns about potential returns on these investments [1][2]. Group 1: Investment Plans - Major tech companies have announced substantial investments in data centers: Meta plans to invest over $600 billion over the next three years, Microsoft $80 billion by 2025, Google $75 billion, and Apple $500 billion over four years, potentially pushing global data center and AI infrastructure investments to over $5 trillion in the next five years [1]. - IBM's CEO Arvind Krishna expressed skepticism about the returns on these investments, stating that the current infrastructure costs make it impossible to achieve returns on the promised multi-trillion dollar investments [2][4]. Group 2: Cost Analysis - Krishna calculated that filling a 1 gigawatt data center costs approximately $80 billion, leading to a total capital expenditure of about $8 trillion if tech companies pursue a total capacity of 100 gigawatts [4]. - He emphasized that this level of investment would require around $800 billion in profits just to cover interest payments, not accounting for depreciation of equipment, particularly AI chips that have a rapid obsolescence rate [4]. Group 3: Comparison to Past Bubbles - Krishna compared the current AI investment frenzy to the internet bubble of the early 2000s, noting that while fiber optics had long-term utility, AI hardware like GPUs has a much shorter lifespan, necessitating expensive updates every five years [5]. - He acknowledged that while some infrastructure can last, the rapid pace of technological advancement in AI hardware raises questions about the sustainability of current investments [5]. Group 4: AGI Potential - Krishna expressed a low probability (0 to 1%) that current technology can achieve Artificial General Intelligence (AGI), contrasting sharply with optimistic statements from other tech leaders [6][8]. - He believes that achieving AGI will require significant advancements beyond current large language models (LLMs) and emphasizes the need for integrating hard knowledge with AI technologies [8]. Group 5: IBM's Strategic Focus - IBM has chosen not to compete in the consumer AI market, focusing instead on enterprise solutions, where it can leverage its long-standing reputation for data protection and reliability [9]. - The company is actively hiring while others in the tech sector are laying off employees, as it aims to enhance productivity through AI tools [9]. Group 6: Quantum Computing Outlook - Krishna predicts that quantum computing could reach practical scale within three to five years, with an estimated market value of $400 billion to $700 billion annually [9]. - He provided a probabilistic timeline for when quantum computing might deliver significant commercial value, suggesting a higher likelihood of breakthroughs within four to five years [10]. Group 7: Industry Perspectives - Krishna's views reflect a broader skepticism within the tech industry regarding the disconnect between current investment levels and realistic return expectations, while still acknowledging the transformative potential of AI for enterprise productivity [11]. - The ongoing debate highlights differing beliefs about the future of AI and AGI, with some companies betting heavily on becoming market leaders through substantial investments [12].

Core Insights - The growth in 2026 is expected to be driven primarily by an AI-driven industrial cycle, with significant capital investments from major tech companies in both China and the U.S. [1] - The market is currently in a phase of striving towards General Artificial Intelligence (AGI), with uncertainties regarding which traditional industries will be disrupted and which new industries will emerge [1] - The Federal Reserve is focused on providing stable financial conditions to support the ongoing tech revolution, which may lead to continued strong performance of tech assets in 2026 [1] Investment Strategy - A long-term perspective and a firm mindset are essential for investing in U.S. and Chinese tech assets [2] - Identifying other high-quality, low-correlation return streams to hedge against tech asset volatility is crucial [2] - The core investment strategy for 2026 may revolve around a combination of technology and Commodity Trading Advisors (CTA) [2] CTA Performance - CTA managers have shown varied performance in 2025, with some achieving returns exceeding 40% due to high volatility in precious metals [3] - CTA funds with neutral strategies have also performed well, with returns over 20% [3] - During significant market pullbacks, certain CTA managers maintained stable performance, highlighting their risk-hedging capabilities [4] Current Market Environment - The current low-interest-rate environment enhances the appeal of CTA's absolute return characteristics [5] - Global factors, including the Fed's continued rate cuts and changing trade dynamics, may trigger commodity market rallies, benefiting CTA products [5] - The AI technology revolution is expected to drive new economic growth, but it may also introduce volatility in capital markets [6][8] Asset Allocation Considerations - Investors should focus on the long-term significance of AI tech assets while also considering strategies to hedge against short-term volatility [8] - The combination of technology and CTA may provide a balanced approach to capturing opportunities in 2026 [9] - Each asset class has unique roles, and a diversified approach may enhance the potential for returns while managing risks [9]

在梵蒂冈这个世界上最古老的权力中心，一群「信徒」正试图警告另一群信徒：末日可能并非来自神罚，而是来自代码。 提到通用人工智能（AGI）引发的末日危机，人们很难第一时间联想到教皇利奥十四世（Pope Leo XIV）。 但谷歌AI研究员约翰-克拉克·莱文（John-Clark Levin）不管这些。 约翰-克拉克·列文是谷歌AI领域的远见者雷·库兹韦尔（Ray Kurzweil）的研究负责人，拥有哈佛大学肯尼迪学院硕士学位和剑桥大学博士学 位。 在库兹韦尔技术公司（Kurzweil Technologies），他负责预测通往ASI的路线图，并重点关注该技术在生命科学领域的应用。 这项研究成果被收录在库兹韦尔去年夏天出版即登上《纽约时报》畅销书榜的著作《奇点更近》（The Singularity Is Nearer）中。 此外，他还是尼尔·弗格森（Niall Ferguson）爵士旗下的宏观战略咨询公司Greenmantle的人工智能高级顾问，定期就人工智能的技术和政策议 题向政府、非政府组织及财富500强企业提供咨询建议。 在过去的十年里，列文在哈佛大学和剑桥大学深入研究人工智能及其影响，并受邀在包括美国海军战 ...

简评：商业不动产REITs试点推出意义重大，将为房企和地方国资提供市场化融资与退出渠道，有效缓 解流动性压力。采取与基础设施REITs并行推进策略，能精准对接商业不动产盘活需求。审核链条简化 有望加速产品扩容，中长期看，有利于盘活万亿级存量资产，降低杠杆，防范风险，为房地产发展新模 式提供金融支持，促进资本市场服务实体经济质效提升。 今年以来，截至12月1日，共有3004只科技创新债券正式发行，发行规模合计达3.18万亿元，发行数量 及总规模相较去年同期分别增长85%和98%，为科技创新企业提供了有力的资金支持。 简评：今年科创债发行明显提速，发行主体及发行规模扩容显著。发行科创债有助于帮助企业融资，为 科创企业提供中长期资金，缓解融资难问题。同时，可以增加债券市场品种，满足多元投资需求，助力 资本市场创新。引导资金流向科技创新领域，提高政策传导效率。 【博时市场点评12月2日】两市震荡调整，成交有所缩量 每日观点 今日沪深三大指数震荡调整，两市成交缩量至1.6万亿。昨日美国供应管理协会（ISM）数据显示，11月 美国制造业PMI从10月的48.7降至48.2，连续第九个月低于50的荣枯线，并创下四个月来的最 ...

Core Concept - Elon Musk has hinted at a new initiative called "Galaxy Mind," aimed at integrating the core capabilities of SpaceX, Tesla, and xAI to deploy solar-powered AI satellites in deep space, thereby overcoming Earth's energy limitations and backing up human knowledge [1][2]. Group 1: Vision and Integration - The vision represents not only a bold technological exploration but also a potential new business model that combines Tesla's energy technology, SpaceX's launch capabilities, and xAI's advanced models into a new space AI solution, potentially opening up significant growth opportunities driven by synergies [2]. - Musk emphasizes that future large-scale AI operations will rely on solar energy from space, stating that to harness a significant portion of solar energy, one must turn to solar-powered AI satellites in deep space, which he sees as the convergence point of the three companies' expertise [3]. Group 2: Company Roles - SpaceX will provide mature rocket launch and spacecraft manufacturing capabilities, responsible for deploying AI satellites into deep space [4]. - Tesla will leverage its expertise in solar and battery technology to provide efficient and sustainable energy solutions for the satellites [5]. - xAI will be tasked with developing cutting-edge AI models capable of large-scale operation on the satellites [6]. Group 3: Business Entity and Trademark - "Galaxy Mind" is not just a technical concept but may evolve into a future business entity, with goals including backing up human knowledge in space as a hedge against the risk of Earth's civilization collapse [9]. - Trademarks for "Galaxy Mind" and "Galactica" have been filed, with indications that "Galactica" could become the official name of the initiative [9]. Group 4: Synergy and Market Potential - Musk's latest move continues his strategy of maximizing synergies within his portfolio, building on the solid foundations of each company [11]. - Tesla's energy division has achieved leadership in efficient solar panels and storage systems, while SpaceX has transformed commercial space and satellite deployment through its Starlink network, and xAI is developing large language models with the potential for AGI [11]. - From an investor's perspective, this integration plan aims to reorganize mature technologies to explore new markets, combining SpaceX's physical transport capabilities, Tesla's energy capabilities, and xAI's intelligent computing to address Earth's energy constraints and create new physical space for AI development [11].

Group 1 - DeepSeek released two official models: DeepSeek-V3.2 and DeepSeek-V3.2-Speciale, with updates available on the official website, app, and API [4] - The cancellation of the VAT exemption for contraceptive drugs and devices will directly impact production and sales companies, requiring them to reassess cost structures and pricing strategies, potentially leading to increased market prices [2] - Vanke has had 570 million yuan of equity frozen for three years, as reported by the Dongguan Intermediate People's Court [2] Group 2 - The movie "Zootopia 2" has become the highest-grossing imported film in mainland China for the year, surpassing 1.9 billion yuan in box office revenue within five days of release [6] - SF Express launched a "late delivery compensation" service, where the company will bear the costs, and couriers will not be held responsible for compensation [6] - Nestlé China confirmed that the merger of its infant nutrition business with Wyeth will not affect existing operations, with the new entity set to launch on January 1, 2026 [7] Group 3 - The Chinese express delivery industry has seen its annual business volume exceed 1.8 billion packages for the first time, reflecting a strong economic momentum [5] - The former CEO of Hema, Hou Yi, announced a new venture called "Lao Cai Rui Xuan," focusing on live-streaming sales [3] - The China CDC reported that the national flu activity has reached a moderate level, with some provinces experiencing high levels of flu activity [4]

Core Viewpoint - The article discusses the limitations of large models in retaining long-term memory, highlighting the challenges faced in practical applications and the need for a more human-like memory system in AI [3][10][25]. Group 1: Limitations of Current AI Models - The large model industry is experiencing a "memory loss" issue, where AI struggles to retain information over extended interactions, leading to repeated questions and irrelevant responses [4][6]. - The technical architecture of current models, such as the Transformer, suffers from attention decay over long sequences, resulting in the loss of earlier instructions during conversations [5][7]. - The lack of a shared memory mechanism among different AI agents leads to fragmented interactions, causing inefficiencies and confusion in customer service scenarios [6][7]. Group 2: Need for Improved Memory Mechanisms - A more sophisticated memory system is essential for AI to evolve beyond simple question-answering capabilities and to develop understanding and reasoning abilities [15][26]. - The concept of memory in AI should not just focus on storing more data but on retaining valuable information that can guide decision-making [11][12]. - The development of a memory infrastructure that allows for shared, manageable, and traceable memory among AI agents is crucial for enhancing their collaborative capabilities [10][22]. Group 3: Redefining AI Memory with "Memory Bear" - The company "Red Bear AI" is working on a product called "Memory Bear," which aims to create a memory system that mimics human memory processes, allowing for better retention and utilization of information [10][28]. - This system includes short-term working memory for task connections and long-term memory for knowledge retention, enabling AI to respond more accurately and contextually [14][18]. - The introduction of a structured memory graph allows for the analysis and retrieval of relevant memories, significantly improving the efficiency and accuracy of AI responses [17][18]. Group 4: Implications for Business and Future of AI - The ability of AI to retain memory will fundamentally change its role in business, allowing it to replace human-like interactions in customer service and other sectors [21][22]. - As AI develops a continuous memory, it will be able to understand user context and history, enhancing trust and effectiveness in various applications [22][26]. - The evolution of memory systems in AI is seen as a critical step towards achieving general artificial intelligence (AGI), where memory plays a vital role in reasoning and learning [26][28].

Core Insights - The integration of intelligent driving technology is reshaping lifestyles at an unprecedented pace, driven by advancements in artificial intelligence and a unique market environment in China [1][3] - The Chinese intelligent driving industry is transitioning from a phase of rapid growth to one of high-quality development, with regulatory frameworks being strengthened alongside pilot programs for higher-level autonomous driving [3][4] - The rapid adoption of electric vehicles is providing an optimal platform for intelligent driving technologies, creating a virtuous cycle between electrification and intelligence [4][6] Industry Trends - The emergence of cognitive intelligence technologies is transforming intelligent driving from a rule-based tool to a cognitive-driven entity, with new architectures like end-to-end and VLA opening new possibilities for high-level autonomous driving [3][5] - The intelligent driving sector is witnessing a clear focus on L4-level scenario-based applications, with significant investments directed towards areas like unmanned delivery and logistics [6][7] - Key components of the supply chain, such as sensor manufacturers and chip companies, are receiving substantial funding, highlighting their foundational role in the development of autonomous driving [7] Regulatory Environment - The regulatory landscape is evolving, with policies being introduced to facilitate the testing and commercialization of L3-level and above autonomous driving technologies in multiple cities [3][4] - The dual approach of relaxing pilot programs while simultaneously enhancing regulatory frameworks is creating clearer competitive advantages for companies with core competencies [3][4] Investment Landscape - Investment activities in the intelligent driving sector are increasingly concentrated in later-stage financing, indicating a shift from technology validation to large-scale commercial applications [7] - Traditional automotive companies are actively participating in investments to address technological gaps, while collaborations within the supply chain are emerging to build ecological advantages [7] Future Outlook - The competition in intelligent driving is entering a new phase where success will depend on the ability to integrate technology, compliance, and commercialization effectively [9] - The industry is at a historical turning point, with the potential for new industry giants to emerge from the convergence of technology, policy, and market dynamics [8][9]

Core Insights - The global AI sector experienced significant developments in November, including increased investments, advancements in AI models, and innovative approaches to data center operations in space [1] Investment Trends - Russian President Putin emphasized the importance of AI for national sovereignty and proposed a comprehensive plan for developing AI technologies in Russia, including the establishment of a leadership department for AI affairs [2] - U.S. President Trump initiated the "Genesis Task" to create a comprehensive AI platform aimed at enhancing national security and productivity, likening its ambition to the Manhattan Project [2] - Microsoft announced a total investment of $15.2 billion in AI projects in the UAE, with over $7.3 billion planned for 2023 and an additional $7.9 billion from 2026 to 2029 [3] - Amazon Web Services (AWS) and OpenAI entered a strategic partnership worth $38 billion to provide cloud computing infrastructure for AI workloads over the next seven years [3] Innovations in AI Infrastructure - Several tech companies are exploring the concept of relocating data centers to space to meet rising computational and energy demands, leveraging solar power [4] - The U.S. company Nebula successfully launched the "Nebula-1" satellite, which includes an NVIDIA GPU for testing high-performance AI computing in space [4] - Google unveiled its "Solar Catcher" project, aiming to create a space-based machine learning data center powered by a network of interconnected satellites [4][5] Advancements in AI Models - OpenAI released the GPT-5.1 series, featuring an "Instant" version for general users and a "Thinking" version for advanced reasoning tasks [6] - Elon Musk's xAI introduced the Grok 4.1 model, enhancing creative and emotional interaction capabilities [6] - Google launched the Gemini 3 model, touted as its most powerful AI agent to date, advancing towards general artificial intelligence (AGI) [6] - Chinese AI firm DeepMind unveiled the DeepSeek-Math-V2 model, achieving gold medal-level performance in international mathematics competitions [6]