每经编辑｜金冥羽 宋思艰 据央视财经援引路透社1月31日报道，美国太空探索技术公司也就是SpaceX，于1月30日向美国联邦通信委员会提交申请，计划发射多达100万颗卫星，以 打造"轨道数据中心"。与此同时，近期频繁传出消息称SpaceX公司正考虑IPO。市场正密切关注企业家马斯克推进太空算力布局的最新动态。 SpaceX提交给美国联邦通信委员会的文件中显示，SpaceX申请发射并运营一个由多达100万颗卫星组成的卫星网络。根据SpaceX的描述，这将是一个拥有 前所未有计算能力的"轨道数据中心系统"，可为先进的AI模型及相关应用提供支持。SpaceX认为，相比于在地球上建设传统的AI算力数据中心，利用太 空中的太阳能，运营或维护成本极低，借此可实现能源效率的革命性提升。 马斯克在社交平台X上转发了关于SpaceX申请发射多达100万颗卫星的报道。 据知情人士向美国彭博新闻社披露，马斯克正考虑合并太空探索技术公司（SpaceX）和他旗下人工智能企业xAI公司的可能性，或是合并太空探索技术公 司和特斯拉公司。 报道称，马斯克在太空部署人工智能数据中心的设想将可能通过上述不同路径实现。如果工程技术允许，其中一个方 ...

文｜邓咏仪 编辑｜苏建勋 1月，一个疯狂的模型大更新季度刚刚过去，刚刚发布新模型K2.5的Kimi，来到一个关键节点。 《智能涌现》获悉，近期Kimi在和投资人的沟通中表示，公司的海外收入已超过国内收入，新模型K2.5发布后，全球付费用户已有4倍增长。 这一变化恰好发生在新一代模型K2.5发布后的短短几天内。 继上一代模型K2发布会后，K2.5继续引发了海外热潮。在Openrouter上，K2.5的排名已经来到第三位，仅次于Claude Sonnet 4.5和Gemini 3 Flash。 来源：OpenRouter 事实上，前一代模型K2发布后，Kimi从10月开始商业化，进程已经算很快。 2025年末的内部信中，杨植麟提到，2025年11月以来，Kimi的海外API收入增长4倍。海外和国内付费用户数月度环比增速超过170%。 1月27日，Kimi发布并开源K2.5模型，这是Kimi迄今最智能的模型，采用原生多模态架构，能力覆盖视觉理解、代码生成、Agent集群、思考与非思考模 式。 在HLE（人类最后考试）、BrowseComp、SWE-Bench Verified等基准测试中，K2.5都能达到开源SOT ...

格隆汇2月2日｜甲骨文公司当地时间周日宣布，计划在2026年通过债务和股权融资相结合的方式筹集 450亿至500亿美元，以扩建其云基础设施产能。该公司在一份声明中表示："甲骨文正在筹集资金以扩 充产能，以满足我们最大的云基础设施客户的合同需求，这些客户包括 AMD、Meta、NVIDIA、 OpenAI、TikTok、xAI 等。"该公司计划通过股权挂钩工具和普通股发行（包括强制可转换优先股）以 及一项规模至高达200亿美元的ATM股票发行计划，筹集目标金额的一半资金。甲骨文表示，计划于 2026年初通过单次债券发行筹集剩余资金。 来源：格隆汇APP ...

这项合作一度被视为双赢：英伟达能向 OpenAI 出售大量顶级 AI 芯片，进一步扩大数据中心业务版图；OpenAI 则能获得扩张 AI 基础设施所需的资金支 持。 最新情况显示，合作正在遇到阻力。上周五，《华尔街日报》报道称，英伟达内部有人开始质疑交易，两家公司的谈判并未走出初期阶段。报道称，双方正 在重新评估合作前景，最新讨论可能涉及在 OpenAI 当前融资轮中进行规模达数百亿美元的股权投资。 报道还提到，黄仁勋曾批评 OpenAI 的商业策略"缺乏纪律"，并对 OpenAI 在与谷歌、Anthropic 竞争中的处境表示担忧。 IT之家 2 月 2 日消息，当地时间 2 月 1 日，据彭博社报道，英伟达 CEO 黄仁勋明确表态，公司并没有必须投资 OpenAI 的责任，而是更愿意"走一步看一 步"，逐步投入。 此前，OpenAI 曾暗示英伟达可能最高投资 1000 亿美元（IT之家注：现汇率约合 6965.73 亿元人民币），黄仁勋对此表态称，这个数字从来不是正式承诺， 更不意味着英伟达会一次性砸下 1000 亿美元。"OpenAI 确实邀请我们最多可以投到 1000 亿美元，我们很高兴也很荣幸，但投 ...

当黄金白银开始像狗狗币一样波动的时候，你我都应该承认，全球资产正在快速MEME化。 而这种"液化"不仅仅发生在金融市场。 如果你把目光移向我们赖以生存的另一个领域——信息与知识，你会发现一场惊人相似的"去锚化"运动正在发生。这就是GEO的诞生。 这里的"Meme化"，指的并不是那几只印着柴犬或青蛙表情包的加密代币重新火热，而是指那些曾在过去几千年里被人类视为文明基石、信贷锚点、最终 避险地的一般等价物——黄金和白银，正在以一种难以理解的速度上下波动。 在刚刚过去的1月中，黄金在上涨25%后，一夜大跌最多12.85%，创出40年以来的最大跌幅；而白银的风格则更是飘逸，从1月初的71美元上涨至121美元 后，一夜最低跌到了74美元，日跌幅最高36.33%。 传统的宏观分析框架：那些关于实际利率、通胀预期、地缘政治的分析似乎正在失效。这些原本应该稳定的资产，现在却像风中的羽毛一样，随着社交媒 体的情绪脉冲和量化算法的资金流向剧烈震荡。 这就是"资产Meme化"的本质：权重的丧失。 在这个时代，不再有"核心资产"与"边缘资产"的区别，所有资产都在被同一个逻辑重构：流动性的暴力与情绪的共振。 当避险资产本身开始表现出M ...

Group 1 - Nvidia's CEO Jensen Huang stated that the proposed $100 billion investment in OpenAI was "never a commitment" and emphasized a cautious approach to investment [1][3] - The initial letter of intent signed in September indicated Nvidia's intention to invest up to $100 billion to support OpenAI in building data centers and AI infrastructure using Nvidia chips [2] - Huang expressed concerns about OpenAI's business discipline and competition, which contributed to doubts within Nvidia regarding the investment [3] Group 2 - Huang affirmed his belief in OpenAI's work, calling it "incredible" and one of the most consequential companies of the time, while stating that Nvidia would still invest significantly [4] - OpenAI is reportedly accelerating plans for a potential initial public offering (IPO) that could occur as soon as the fourth quarter [4] - The investment landscape is being closely monitored as it will test investor appetite for large-scale AI growth stories that require substantial spending on chips and infrastructure [5] Group 3 - Recent research indicates that OpenAI's ChatGPT leads in consumer usage at 83%, significantly ahead of Google's and Microsoft's chatbots at 48% and 30% respectively [6] - The data reflects initial exposure to AI platforms rather than sustained engagement, suggesting that the first platform users encounter often becomes the reference point for the category [7]

智通财经APP获悉，英伟达(NVDA.US)首席执行官黄仁勋表示，公司拟议的对OpenAI最高1000亿美元 投资"从来不是一项承诺"，公司将"逐轮"考虑任何融资。他在上周日接受采访时表示："那从来都不是 一项承诺。他们邀请我们最多投资1000亿美元，当然，我们对此感到非常高兴，也很荣幸能收到这样的 邀请，但我们会一步一步地进行投资。" 软银也正就追加高达300亿美元投资事宜与OpenAI进行磋商，相关磋商仍处于早期阶段，最终投资金额 及具体条款仍可能发生变动。此前，软银已通过愿景基金及联合投资等方式，向OpenAI投资了逾400亿 美元。 英伟达计划投资于其先进AI芯片的关键买家OpenAI，这一举动在过去一年引发了人们对AI交易循环性 质的担忧。投资者日益质疑这些合作伙伴关系(即科技公司投资于购买其产品的AI企业)是否在人为支撑 需求。在另一项类似交易中，英伟达近期宣布计划向云计算提供商CoreWeave(CRWV.US)追加投资20亿 美元，该公司同样是其重要客户。 OpenAl正进行新一轮融资 多方将参与 近日有消息称，OpenAI正在进行新一轮融资，最高融资额达到1000亿美元，该轮融资或将使其估值 ...

Core Insights - Cybersecurity is increasingly recognized as integral to national security, especially with the rapid evolution of generative AI, which presents new security challenges [1] - The essence of AI competition among nations is not merely technological breakthroughs but the efficient scaling of AI applications across ecosystems, with safety and trust being critical barriers to adoption [2] Group 1: AI Security Challenges - Generative AI has inherent flaws that lead to various security challenges, including data pollution, content violations, and privacy breaches, which surpass traditional security concerns [1][2] - The foundational digital architecture of AI systems cannot completely eliminate vulnerabilities, leading to uncertain safety boundaries [2][3] Group 2: Proposed Solutions and Frameworks - A new theory of intrinsic AI safety is proposed, suggesting that safety should be built into AI systems from the ground up rather than relying on post-hoc fixes [3] - The establishment of credible safety assessment standards in AI, akin to those in the automotive industry, is recommended to guide government regulation and corporate self-discipline [3][4] Group 3: Legislative and Regulatory Developments - The first fifth-level information system registration certificate was awarded, and a white paper on generative AI cybersecurity was released, outlining risks and implementation paths for security measures [4] - There is a pressing need for legislative action on cybersecurity regulations to enhance the existing legal framework and mechanisms for network security protection [5]

Core Viewpoint - Investors are expressing concerns regarding Microsoft's AI-driven growth strategy, leading to a significant drop in its stock price by over 7% following the release of its fiscal 2026 second-quarter earnings [1] Group 1: Azure Performance - Revenue for Microsoft's Azure and other cloud services increased by 39% in the quarter ending December 31, slightly below Wall Street's expectations [2] - CFO Amy Hood indicated that Azure's growth could have exceeded 40% if all available GPUs had been allocated to cloud services instead of being used for first-party applications like Microsoft 365 Copilot and GitHub Copilot [3] Group 2: Strategic Decisions - CEO Satya Nadella emphasized a long-term strategy by allocating supply-constrained chips to areas that enhance customer lifetime value, although this approach has not resonated well with investors [4] Group 3: OpenAI Concerns - Microsoft's remaining performance obligations reached $625 billion by December 31, with 45% linked to OpenAI's expansion initiatives, raising concerns about the sustainability of this investment given OpenAI's projected losses of $14 billion in 2026 [5] - The increasing cash burn of OpenAI has led investors to question whether Microsoft will realize its expected future revenue, especially if OpenAI struggles to meet its capital spending needs [6]

Core Viewpoint - Microsoft experienced its largest single-day dollar loss in history, with a decline of $357 billion in market value despite reporting strong earnings for Q2 of fiscal 2026, raising questions about investor confidence and future growth prospects [1]. Financial Performance - Revenue for the quarter increased by 17% to $81.3 billion, with adjusted earnings reaching $4.14 per share, surpassing the consensus estimate of $3.91 [1]. - Operating margin stood at 47.1%, and Microsoft Cloud revenue exceeded $50 billion for the first time [1]. Market Reaction - Microsoft shares fell by as much as 12% during intraday trading, closing down 10% at $433.50, marking the seventh-largest percentage decline since the company went public in 1986 [1]. - The stock's performance post-earnings report was notably stagnant, indicating persistent investor concerns [1]. Growth Concerns - Azure cloud platform growth was reported at 38% in constant currency, but fell short of Wall Street's whisper number of 39.4%, contributing to market unease [1]. - Capital expenditures surged to $37.5 billion, a 66% increase year-over-year, highlighting the competitive pressures in the AI and cloud sectors [1]. AI Business Insights - Microsoft 365 Copilot, the AI assistant integrated into Office apps, has 15 million paid users, which is only about 3% of the 450 million paid seats in Microsoft 365 [1]. - Concerns were raised regarding 45% of Microsoft's $625 billion in remaining performance obligations being tied to OpenAI, with $281 billion of that backlog committed to a single customer [1]. Analyst Perspectives - Analysts expressed skepticism about Microsoft's ability to prove the value of its investments in AI, despite a 16% increase in Microsoft 365 commercial revenue [1]. - Some analysts maintained a positive outlook, with Morningstar keeping a $600 fair value estimate and noting that demand for AI and cloud services continues to outpace supply [1]. - Other analysts, like Wedbush, acknowledged the friction between long-term investments and short-term investor expectations, suggesting that 2026 could be a pivotal year for Microsoft [1].