Z Highlights 巨大的TAM扩张 上个月，我们解释了为何某些公司会成为"模型破坏者"（modelbusters）——或者说公司的增长速度和持续时间比任何人都要快很多。 当创始人构建了两样东西时，模型破坏者就会出现：一是突破性产品，释放出远超任何人预测的市场规模；二是抓住这一风口的商业模式。尽管模型破坏 者可能偶然出现，但最大、最优秀的那些往往在平台转换期蓬勃发展，此时新的技术基础让创始人能够重新定义产品和商业模式。 这正是AI目前所处的位置，但这一机会的规模可能远超我们迄今为止所见的一切。尽管我们仍处于早期阶段，但AI已经在以十分之一的成本提供十倍的产 品体验，并且我们开始看到新的商业模式出现以捕获这一价值。如果您使用SaaS或移动时代的可比数据来预测AI的规模，比如在ChatGPT出现时，您的模 型早已严重低估了AI的增长。 AI已在以十分之一的成本提供十倍的产品体验，同时新商业模式开始涌现以捕获这一价值。 AI公司有能力进攻6万亿美元的白领服务市场，规模是美国企业软件支出的20倍。 过去三年，智能成本每年下降超10倍，而模型能力约每7个月翻一番。 图片来源： a16z 在我们看来，很难高估AI会 ...

Core Viewpoint - Microsoft has removed traditional methods for creating local accounts during the Windows 11 setup process, requiring users to log in with a Microsoft Online Account to complete the installation [1][3][12]. Group 1: Changes in Installation Process - The latest updates (Build 26120.6772 and 26220.6772) have blocked previous workarounds for offline installation, such as using command prompts to bypass the online account requirement [1][7]. - Users are now forced to connect to the internet and log in to a Microsoft account during the initial setup, which impacts the installation experience [3][12]. Group 2: User Concerns and Motivations - Users previously opted for offline installation to avoid the risk of losing the 7-day return policy once the Windows system was activated online [9][12]. - Privacy concerns also drove users to prefer local accounts, as data stored locally minimizes the risk of personal information being exposed through cloud synchronization [11][12]. - Many users sought to avoid advertisements and unnecessary features that come with using a Microsoft account, which integrates them into Microsoft's ecosystem [11][12]. Group 3: Microsoft's Business Strategy - The push for Microsoft account usage is driven by commercial interests, allowing Microsoft to collect user data for targeted advertising and increase subscription rates for services like OneDrive and Office 365 [13]. - Tying accounts to the Microsoft Store facilitates the management of paid applications, ensuring revenue generation for the company [13].

来源：环球网 而微软、谷歌和苹果在接到置评请求后，也未立即给出回应。OpenAI发言人则引导关注其此前发布的 博客文章，文章中强调了人工智能竞争过程中存在的诸多问题，尤其是在输入和分销层面遭遇的瓶颈， 其中对谷歌的批评尤为明显。 彭博社称，尽管目前OpenAI的警告尚未升级为正式的反垄断投诉，但这一举动无疑已经为科技行业敲 响了警钟。在人工智能推广的关键时期，以及OpenAI与微软重新调整合作关系的大背景下，一旦矛盾 升级，或将在科技巨头之间引发激烈的紧张关系。 此前，微软与OpenAI携手后，迅速将人工智能技术融入自身产品线。但随着业务的发展，两家公司之 间的竞争也日益白热化。微软的个人Copilot应用与OpenAI的ChatGPT在市场上正面交锋，而OpenAI在 企业市场的不断拓展，也直接冲击了微软的主要收入来源。（旺旺） 此前，OpenAI代表前往欧盟反垄断负责人特雷莎·里贝拉办公室，详细阐述了在竞争过程中遭遇的诸 多"困难"。OpenAI指出，在与这些老牌科技企业同场竞技时，面临着巨大的竞争压力，大型平台存在 锁定客户的潜在风险，这严重影响了市场的公平竞争环境。 彭博社报道指出，一位知情人士透露， ...

【环球网报道 记者 索炎琦】据英国《卫报》、路透社等媒体报道，英国前首相苏纳克已加入美国微软 公司和美国人工智能（AI）初创企业Anthropic公司担任高级顾问。《卫报》称，苏纳克也成为继英国 前副首相克莱格之后担任硅谷巨头有偿顾问的英国前高官。 同样报道相关消息的法新社称，ACOBA已在建议信中告知当前仍是英国议会议员的苏纳克不要代表上 述公司游说政府。 据介绍，苏纳克曾于2022年10月至2024年7月担任英国首相。法新社说，此前有报道称，苏纳克在去年 大选中落败后，正考虑前往硅谷工作。 《卫报》称，"英国商业任命咨询委员会（ACOBA）"当地时间周四（9日）发布了上述消息。路透社 称，苏纳克本人同日在美国职业社交网站"领英"上发帖证实新职，并表示这些职位的收入将全部捐赠给 他与妻子穆尔蒂共同创办的一家名为"里士满计划"的慈善机构。 路透社介绍，苏纳克将为Anthropic公司提供战略、宏观经济和地缘政治趋势方面的建议；在微软，苏 纳克将就宏观经济和地缘政治趋势提供战略视角。 ...

Market Overview - The US stock market experienced increased volatility due to a lack of key economic data, with investors expressing concerns over potential bubble risks in previously high-performing tech stocks and cyclical assets, leading to significant profit-taking pressure [1] - Despite a late-session optimistic forecast regarding Microsoft’s data center demand, major indices such as the S&P 500, Nasdaq, and Dow closed lower [1] - There is a noticeable sector rotation, with investors selling off economically sensitive sectors and shifting towards defensive sectors like healthcare and consumer staples, reflecting a hedge against macroeconomic uncertainty [1] Nvidia Developments - Nvidia continued its upward trend, reaching both intraday and closing historical highs, with a market capitalization nearing $4.7 trillion [3] - The US Department of Commerce has granted Nvidia an export license to sell chips worth billions to the United Arab Emirates, in line with a bilateral AI agreement established in May [3] Oracle's Performance - Oracle shares rose by 3.06%, driven by positive analyst ratings, with TD Cowen designating it as a "Top Pick" and Baird initiating coverage with an "Outperform" rating, highlighting strong growth potential in its AI business [4] Microsoft’s Supply Challenges - Microsoft predicts that the computing power shortage crisis will persist until 2026, with significant physical space and server shortages in key US data center regions, directly impacting new subscriptions for Azure cloud services [5] - The current supply constraints are more prolonged than previously indicated, with Azure expected to generate over $75 billion in revenue for FY2025, making the supply-demand imbalance particularly concerning [5][8] Dual Demand Impact - The capacity crisis is driven by two core demands: strong computational needs from AI and traditional cloud infrastructure, with Microsoft’s CTO noting that the demand for AI workloads has outpaced capacity expansion efforts [8] - Traditional workloads, including those from Microsoft’s Office suite and clients like OpenAI, are also contributing to the demand for computing resources [8] Client Strategies - Clients facing Azure capacity limitations are adopting pragmatic strategies, such as shifting to other cloud providers like AWS and Google Cloud, or utilizing multiple Azure regions while awaiting additional capacity [9] Microsoft’s Remedial Actions - Microsoft is actively seeking to balance supply and demand, with a spokesperson indicating that most Azure services still have available capacity for existing customers [10] - The company is guiding clients to other regions when preferred facilities are full, although this may increase system complexity and data transfer delays [11] Infrastructure Expansion and Supply Chain Issues - Microsoft has initiated a historic data center construction surge, adding over 2 gigawatts of capacity in the past year, yet faces significant delays due to supply chain bottlenecks affecting critical components [12] - The construction timeline for data centers can extend for years, complicating rapid expansion efforts amid ongoing global supply chain challenges [12]

Market Overview - US stock indices closed lower on Thursday, with the Dow Jones Industrial Average recording its largest single-day drop in a month, down 243.36 points to 46358.42, a decline of 0.52% [1] - The S&P 500 index fell 18.61 points to 6735.11, down 0.28%, while the Nasdaq Composite dropped 18.75 points to 23024.63, a decrease of 0.08% [1] - Among the 11 sectors of the S&P 500, the materials sector saw the largest decline, while the consumer staples sector was the only one to rise [1] Earnings Season Insights - The upcoming earnings season is anticipated to show whether corporate profits can continue the stable growth seen in the past two quarters, with analysts expecting an 8.8% year-over-year increase in S&P 500 earnings for Q3, down from 13.8% in the previous quarter [2] - Major banks including JPMorgan Chase, Goldman Sachs, Citigroup, and Wells Fargo are set to report their earnings next week [2] Economic Indicators - The yield on the US 10-year Treasury rose by 1.7 basis points to 4.15%, while the 2-year Treasury yield increased by 1.3 basis points to 3.60% [3] - The futures market indicates a 94.6% probability of a 25 basis point rate cut by the Federal Reserve in October, as the job market may weaken further [2] Commodity Prices - International oil prices fell, with WTI crude oil futures down 1.66% to $61.51 per barrel, and Brent crude down 1.55% to $65.22 per barrel [3] - Gold prices also retreated, with spot gold down 1.62% to $3976.76 per ounce, and COMEX gold futures down 1.96% to $3990.9 per ounce [3] Corporate Developments - Delta Air Lines reported stronger-than-expected Q3 earnings and raised its guidance for the current quarter, resulting in a 4.3% increase in its stock price [2] - The Keator Group's partner noted that the market is adjusting due to a lack of new economic data and clear policy signals, which is a natural occurrence [2]

Core Viewpoint - The current market environment, characterized by soaring stock prices of tech giants like Nvidia and Microsoft amid the AI wave, raises concerns reminiscent of the "irrational exuberance" warning from former Fed Chairman Alan Greenspan in 1996, with financial leaders warning of significant market correction risks in the next six months to two years [1] Valuation Metrics - The S&P 500's price-to-earnings (P/E) ratio is approximately 23 times, nearing a five-year high and significantly above the ten-year average of 18.7 times, yet still below the peak of around 25 times during the dot-com bubble of 1999-2000 [1] - The tech sector's P/E ratio stands at about 30 times, exceeding the long-term average of 21.4 times, but remains far from the extraordinary level of 48 times seen during the dot-com bubble [1] Market Sentiment - The trading volume of stock options, a measure of bullish sentiment, is approaching a four-year high, heavily concentrated in AI-related tech stocks, indicating strong market enthusiasm for tech stock gains [1] Analyst Perspectives - Despite market concerns, many analysts and investors believe the current situation differs from 2000, with Goldman Sachs analysts asserting that the current rally is driven by "fundamental growth rather than irrational speculation," highlighting that leaders in the AI field are established companies like Microsoft and Google with solid business models and substantial profits, unlike many unprofitable startups of the past [1] Investor Positioning - Data shows that institutional investors are maintaining a neutral position, while retail investors are allocating more funds to bonds and money markets, indicating a lack of extreme enthusiasm in the market [1]

Market Overview - Major US indices experienced declines, with the Dow Jones Industrial Average down by 243.36 points (-0.52%) to 46358.42, and the S&P 500 down by 18.61 points (-0.28%) to 6735.11 [2] - European markets also saw losses, with the Euro Stoxx 50 down by 24.17 points (-0.43%) to 5625.56, while the German DAX gained slightly by 14.12 points (0.06%) to 24611.25 [2] - Asian markets showed mixed results, with the Nikkei 225 up by 845.45 points (1.77%) to 48580.44, while the Hang Seng Index fell by 76.87 points (-0.29%) to 26752.59 [2] Economic Indicators - The US Bureau of Labor Statistics (BLS) is expected to release the September CPI data, which is crucial for determining the cost-of-living adjustments for Social Security benefits and may influence the Federal Reserve's interest rate decisions [2] - Federal Reserve Governor Barr indicated that the long-term neutral interest rate has slightly increased, but the change is not significant [3] - Following the first rate cut in September, Barr emphasized the need for caution in further policy adjustments due to persistent high inflation, which is expected to remain above target until the end of 2027 [4] International Developments - The US Treasury intervened in the Argentine foreign exchange market, purchasing Argentine pesos and establishing a $20 billion currency swap framework [4] - The Israeli government approved a ceasefire agreement in Gaza, which includes the withdrawal of Israeli forces and the release of hostages by Hamas [5] - The US and Saudi Arabia are making progress on a semiconductor export agreement, with potential deals involving major US chip companies [5] Corporate News - Tesla is facing a safety investigation regarding its Full Self-Driving (FSD) system, linked to 44 reported accidents involving traffic violations [6] - Microsoft forecasts that the data center resource shortage will persist until mid-2026, affecting Azure subscription availability [6] - Google Cloud launched the Gemini enterprise version, priced at $30 per month, to compete with Microsoft and OpenAI in the enterprise AI market [6] - OpenAI's CEO announced plans for more significant infrastructure deals, with total agreements this year reaching $1 trillion [6] - Intel revealed details about its first 18A process PC chip architecture, expected to begin shipping later this year and widely available by January 2026 [7]

Core Insights - Microsoft is facing a prolonged data center supply shortage, likely extending until 2026, which is longer than previously anticipated by management [1][4] - The demand for AI computing power continues to surge, with global AI startups raising a record $192.7 billion, indicating a fierce competition in the AI sector [1] - Major cloud providers, including Microsoft, Amazon, and Google, are struggling with supply constraints for high-performance AI servers and traditional cloud computing servers [2][3] Group 1: Data Center Supply Constraints - Microsoft’s internal forecasts indicate that new Azure subscriptions in key regions like Northern Virginia and Texas will be limited until mid-2026 due to severe shortages [2][4] - The shortage affects both AI GPU-based servers and traditional CPU-based cloud computing services, impacting Microsoft's Azure platform, which is projected to generate over $75 billion in revenue for FY2025 [3][6] - The ongoing supply constraints are attributed to the rapid expansion of AI infrastructure investments, which are driving the demand for new data centers [3][9] Group 2: Competitive Landscape and Market Dynamics - Microsoft, along with its competitors, has consistently reported an inability to meet all customer demands for cloud computing and AI capabilities over the past six quarters [6] - The AI infrastructure investment wave is expected to reach $2 trillion to $3 trillion, driven by unprecedented demand for AI computing resources [10] - Analysts have raised their forecasts for AI infrastructure spending among major tech companies, projecting a rise from $420 billion in 2026 to $490 billion [14]

Group 1 - The most profitable business for Yingshi Juifeng is e-commerce, with a single T-shirt selling 200,000 units this year, shifting focus from traditional video production and content creation [3][5] - The revenue share from TVC advertising has decreased to around 10% from being the most profitable segment previously [3] - Tim Pan Tianhong, the founder, emphasizes the significant scale effect in e-commerce, where user-driven product promotion leads to high sales [3] Group 2 - Xiaopeng Motors has appointed Liu Xianming, head of the world base model, as the new leader of its autonomous driving center, replacing Li Liyun [8] - OPPO has made personnel adjustments, with Liu Zuohua overseeing overseas markets and Duan Yaohui managing the headquarters marketing team [9] - Alibaba has entered the embodied intelligence sector, forming a team led by the head of its Qwen large language model technology [10] Group 3 - The U.S. Walmart has listed the Yushu G1 humanoid robot at a price of $21,600, reflecting a 55% premium compared to its price in China [10][13] - The G1 robot is becoming a preferred experimental platform for AI and robotics developers due to its competitive pricing and performance [13] - Yushu Technology is preparing for a potential IPO, aiming to expand its sales channels and developer ecosystem [13] Group 4 - Li Auto's MEGA model delivered 3,271 units in September, accounting for nearly 10% of the brand's total deliveries for the month [21][22] - The overall delivery of Li Auto vehicles reached 33,951 units in September, with a total of 143,102 units delivered historically [21][22] Group 5 - Xiaopeng Motors plans to announce significant breakthroughs in physical AI at its upcoming AI Technology Day, focusing on advancements in its world base model [23] - The company has been developing its physical AI base model for over a year, aiming to enhance its capabilities in autonomous driving [23] Group 6 - Apple CEO Tim Cook is reportedly stepping down, with John Ternus, the senior vice president of hardware engineering, being the leading candidate for succession [26][27] - Volkswagen is restructuring its software department, with the Cariad team nearly disbanded, aiming to enhance its competitiveness against Tesla and Chinese EV makers [27] Group 7 - Google has tightened its remote work policy, limiting employees to a maximum of four weeks of remote work per year, with any remote work day counting as a full week [28] - Microsoft is migrating GitHub's infrastructure to its Azure cloud platform over the next two years, aiming to address capacity challenges and support AI growth [29]