2025.12. 04 本文字数：2885，阅读时长大约5分钟 作者 | 第一财经 后歆桐 标普500指数重新逼近历史新高。但数据显示，美股科技股在本轮美股行情中未能一如往常成为领涨 板块，市场对人工智能（AI）泡沫的担忧导致AI概念龙头股英伟达、微软等股价受到拖累。 相比之下，自10月28日以来，标普500信息技术指数同期反而整体下跌了4.2%，"美股科技股七巨 头"（下称"美股七巨头"）的股票整体下跌1.3%，仅谷歌母公司Alphabet仍录得18.1%的涨幅。AI 领军股英伟达跌去9.7%，AI领军企业股整体下跌8.4%，AI软件股整体下跌8.5%。 而在4月美国总统特朗普宣布关税政策导致大幅抛售后，科技股在4月8日~10月28日期间始终领涨标 普500指数的上一轮反弹。"美股七巨头"期间上涨69.8%，英伟达期间上涨108.8%。AI领军企业股 整体上涨124.2%，AI软件股整体上涨67.9%。10月29日，科技板块在标普500指数中的权重还一 度创下历史新高，达到约36%。 这被市场人士解读为，由于对股票高估值以及在AI算力、基础建设方面的巨额支出能否换来切实利润 的担忧越发盛行，美股投资者对AI ...

AI的支出热潮如此之大，以至于它已经成为宏观经济故事本身。 标普500指数重新逼近历史新高。但数据显示，美股科技股在本轮美股行情中未能一如往常成为领涨板块，市场对 人工智能（AI）泡沫的担忧导致AI概念龙头股英伟达、微软等股价受到拖累。 但随着科技股、AI概念股走势落后，华尔街两大机构——贝莱德（BlackRock）和美国银行（Bank of America）近 期都开始反驳"AI泡沫"论，并表示本轮AI热潮是由真正的企业投资、收益和生产率增长推动的，而非21世纪00年 代互联网泡沫那样的非理性繁荣。 贝莱德投资研究所所长博威（Jean Boivin）称，鉴于AI继续以"前所未有"的规模和速度扩张，将AI热潮描述为泡 沫是"不完整的"。 而在4月美国总统特朗普宣布关税政策导致大幅抛售后，科技股在4月8日~10月28日期间始终领涨标普500指数的 上一轮反弹。"美股七巨头"期间上涨69.8%，英伟达期间上涨108.8%。AI领军企业股整体上涨124.2%，AI软件股 整体上涨67.9%。10月29日，科技板块在标普500指数中的权重还一度创下历史新高，达到约36%。 这被市场人士解读为，由于对股票高估值以及在A ...

以下文章来源于新皮层NewNewThing ，作者陆彦君 新皮层NewNewThing . 关注AI，提供洞察。 2025.12. 04 本文字数：3395，阅读时长大约6分钟 记者 | 陆彦君 数据整理 ：陆彦君｜ 作图 ：程星 编辑 | 王杰夫 过去数周，英伟达股价经历了一轮高位回调与震荡。 目前市值 相 较 于 一 个 月 前 的 高 点 已 下 降 15.4%。虽然一周前最新季度财报发布后，公司超预期的业绩表现一定程度上稳定了市场信心，但隔 天股价的下跌反映着情绪底色依然是消极的。 关键的是，这已不仅仅是英伟达一家公司的悲欢，它的表现是观察AI行业的核心风向标。再加上 Alphabet、微软、亚马逊、Meta、苹果、特斯拉，这七家技术公司继承了商业历史上石油七巨头的 强势，利益捆绑关系紧密，以惊人的市值集中度成为美股的主心骨。 根据中金公司数据，自2022年 年底ChatGPT发布以来，美股「七姐妹」（Magnificent Seven）最高上涨283%，大幅跑赢同期标普 500指数扣除M7后的69%。 最近一周，美国投资人Michael Burry公开警告AI泡沫，质疑以英伟达为中心的技术公司之间的 ...

贝莱德与美银认为，用"泡沫"框架来评估AI投资并无助益，鉴于AI基础设施建设正以前所未有的规模和速度展开，任何基于历史指标的回溯性评估都 是"不完整的"。当前AI热潮由真实投资和盈利驱动，但市场需警惕一个潜在的"空中气穴"——即巨额资本支出与收入变现之间出现暂时性脱节，这可能 引发短期市场波动。 包括贝莱德和美国银行在内的顶级金融机构认为，纠结于AI是否是"泡沫"已无意义，当前真正的焦点应转向如何应对巨额前期投资与商业变现之间的错 位与滞后，这一阶段可能给市场带来新的动荡。 近日，贝莱德投资研究所负责人Jean Boivin明确表示， 在当前阶段，用"泡沫"框架来评估AI投资并无助益。他认为，鉴于AI基础设施建设正以前所未 有的规模和速度展开，任何基于历史指标的回溯性评估都是"不完整的"。贝莱德指出，AI领域的资本支出规模巨大，其本身已构成一个宏观经济事件。 贝莱德在其展望报告中称，AI相关的资本支出雄心是如此庞大，以至于"微观即宏观"。该机构认为，这一投资规模有望推动美国GDP增长持续突破过去 几十年来主导的2%趋势线。这标志着AI不仅是一项技术革新，更是一个足以重塑整体经济格局的宏观驱动力。 AI的宏大叙 ...

Group 1: Technology Industry Developments - Doubao Mobile Assistant responded to privacy infringement claims, stating that user authorization is required for system-level permissions and that the product is not aimed at ordinary consumers [2] - OpenAI's ChatGPT service experienced interruptions, with approximately 3,000 users reporting issues, and the company has implemented measures to mitigate the problems [5] - Nvidia's CFO refuted "AI bubble" claims, asserting that the global economy is in the early stages of transitioning to AI-required data center infrastructure, with AI investments expected to reach $3 trillion to $4 trillion by the end of 2030 [6] Group 2: Market and Sales Insights - IDC projected that global smartphone shipments will grow by 1.5% to 1.25 billion units this year, with Apple's shipments expected to reach a record high of 247 million units, driven by strong demand for the iPhone 17 series [7][8] - The total number of 5G base stations in China surpassed 4.758 million, with a net increase of 507,000 from the previous year, indicating robust growth in telecommunications infrastructure [9] Group 3: Corporate Announcements and Financials - Moer Technology announced its upcoming IPO on the STAR Market, with the company currently not profitable and set to be included in the Sci-Tech Growth Layer upon listing [4] - Airbus maintained its profit guidance of approximately €7 billion but lowered its delivery target for the A320 series due to quality issues with fuselage panels, adjusting the 2025 delivery target to 790 aircraft from 820 [12] - JD Industrial plans to globally issue approximately 211 million shares, with the share price set between HKD 12.7 and HKD 15.5, and the listing date expected on December 11, 2025 [13]

Group 1 - SoftBank founder Masayoshi Son significantly sold off his Nvidia shares in November, stating he needed funds for investments in OpenAI and other projects, expressing regret over the sale [2][8] - Son believes that if artificial intelligence can create 10% of global GDP in the long term, the investment of trillions of dollars would be justified, countering the argument of an AI bubble [2] - Son's investment history includes notable successes such as a $105 million investment in Yahoo during the internet boom, yielding approximately 100 times return, and a $20 million investment in Alibaba, which peaked at over 3000 times return [4] Group 2 - Son's investment strategy involves making large bets on future technologies, often investing heavily in startups within promising sectors, akin to an "All IN" thematic index approach [5] - A significant drawback of this investment style is a lack of deep understanding of individual companies, which can lead to cash flow issues, as seen when SoftBank reduced its Nvidia holdings after other investments incurred losses [6] - Despite selling Nvidia shares, the investment was still highly profitable, yielding around $5.1 billion, but holding onto the shares could have made Son one of the world's richest individuals again [7]

1.【Anthropic启动IPO筹备工作，最早或于2026年上市】12月3日消息，据报道，人工智能初创公 司Anthropic已聘请律师事务所Wilson Sonsini，开始筹备有史以来规模最大的IPO之一，最早或于 2026年进行。据知情人士透露，该公司正在就一轮私人融资进行谈判，估值将超过3000亿美元。据 多位人士透露，该公司也已与多家大型投行讨论了潜在的IPO计划。不过，相关讨论仍处于初步和非 正式阶段，该公司尚未接近选定IPO承销商。Anthropic发言人表示："对于我们这样规模和收入水平 的公司来说，像上市公司一样有效地运营是相当标准的做法。""我们还没有就何时上市甚至是否上市 做出任何决定，目前也没有任何消息要分享。"（华尔街见闻） 2. 【Mistral AI正式发布Mistral 3模型】12月3日消息，法国人工智能初创公司Mistral AI正式发布 Mistral 3模型。Mistral 3模型包含3个小型密集模型 （14B、8B和3B）以及Mistral Large 3。所有 模型均基于Apache 2.0开源协议发布。（第一财经） 5.【千问接入阿里学习模型】12月3日，阿里千 ...

伴随AI市场繁荣，对于"AI泡沫论"的探讨近几个月来持续升温。当地时间12月2日，英伟达CFO科莱特· 克雷斯（Colette Kress）在瑞银全球技术与AI大会上回应行业竞争，并反驳"AI泡沫论"。 华尔街明星基金经理"木头姐"也坚定传达：不存在AI泡沫。最大的风险可能另有它者。 科技巨头密集发债， 信用市场发酵"AI泡沫" 为了加入人工智能（AI）竞赛，科技巨头密集发债。其中最典型的是甲骨文，该公司近月来已经发行 了数百亿美元债券，加之该公司信用评级低于其他云计算巨头，这使得其信用违约掉期（CDS）成为投 资者对冲"AI泡沫"风险的关键工具。 根据ICE数据服务公司的数据，以纽约尾盘信贷衍生品价格计算，甲骨文债务违约保护成本当地时间12 月2日触及2009年3月以来最高位。 外媒评论称，违约保护成本的不断上升，反映出AI领域的大规模投资与投资者预期看到生产率提高和 企业利润增长之间的存在差距。 "AI泡沫论"讨论度升温 伴随AI市场繁荣，对于"AI泡沫论"的探讨近近日来持续升温。 木头姐称最大风险另有它者 当地时间12月2日，英伟达CFO科莱特·克雷斯（Colette Kress）在瑞银全球技术与AI大 ...

Core Viewpoint - The debate around the existence of an "AI bubble" is intensifying, with discussions focusing on AI's economic contributions and its impact on technology stocks [1]. Group 1: Existence of AI Bubble - The emergence of "AI bubble theory" is attributed to four main reasons: high concentration of holdings, concerns over investment returns, market perception, and doubts about technological pathways [3][4]. - High concentration of AI holdings and profit-taking amid market uncertainties, such as tariffs and Federal Reserve interest rate expectations, are leading to profit realization [3]. - Concerns regarding the sustainability of AI capital expenditures have arisen due to the shift from self-financing to debt financing, impacting cash flow expectations for companies like Oracle and Meta [3]. - Historical context suggests that the current high valuations and growth rates of tech companies are causing unease among investors unfamiliar with the underlying logic [3]. - The divergence between GPU and TPU technologies is influencing performance expectations for companies like NVIDIA, with future growth dependent on which technology path prevails [4]. - Current AI valuations differ significantly from the 1999-2000 internet bubble, with free cash flow yields approximately three times higher and forward P/E ratios about 35% lower than during that period [4]. Group 2: Contribution of AI to Economic Growth - AI-related spending is becoming a key driver of global economic growth, with significant contributions expected in the coming years [5]. - In the U.S., AI spending is projected to contribute approximately 0.4 percentage points to GDP growth by 2026-2027, accounting for about 20% of total growth [5]. - The expected productivity growth from AI in the U.S. is estimated to be between 25 to 35 basis points by 2027 [5]. - In contrast, AI's contribution to GDP growth in China is anticipated to be minimal in 2026, with a contribution of about 10 to 15 basis points by 2027 [5]. - The focus on AI's impact on cost reduction in industries, particularly in the service sector, is expected to become more pronounced by 2026 [6]. Group 3: Impact on Technology Stocks - The fundamental factors surrounding AI are expected to influence stock price performance, with a broader distribution of market leadership anticipated in the U.S. stock market by 2026 [7]. - AI applications are seen as a driving force for U.S. stock strategies, with expectations of positive operational leverage effects [7]. - The health of the upstream computing power industry is expected to remain strong, providing opportunities for AI-related tech stocks in the first half of 2026 [8]. - Increased electricity demand from AI may become a focal point for U.S. stocks, with semiconductor and cloud service providers likely to benefit from AI spending [8]. - The Chinese stock market is expected to see significant performance from technology stocks due to advancements in AI, supported by a large domestic market and favorable policies [8].

软银在人工智能领域的投资布局正在加速。除对OpenAI的巨额投资外，公司还以54亿美元收购瑞士 ABB机器人业务，并通过子公司完成对美国半导体设计公司Ampere Computing的全资收购。据悉，软 银已同意在12月底前向OpenAI投资300亿美元，规模超过此前任何投资者对OpenAI的投资。 软银集团创始人孙正义在12月1日东京FII Priority Asia论坛上首次公开回应清仓英伟达一事。孙正义表 示，如果软银拥有"无限资金"支持人工智能投资，他绝不会出售英伟达股票。"卖英伟达的时候我都哭 了"，孙正义坦言。 据软银11月11日发布的业绩报告显示，公司持有的3210万股英伟达股票已于10月以58.3亿美元价格全部 售出。孙正义解释称，此举纯粹是为筹集资金投资OpenAI等人工智能项目。软银与OpenAI签署修订协 议，承诺追加投资225亿美元，将通过愿景基金2号完成全部投资。 孙正义强烈驳斥了外界关于"AI泡沫论"的质疑。他认为持此观点的人"不够聪明"，并指出如果人工智能 未来能贡献全球GDP的10%，即便投入数万亿美元也远不算泡沫。"泡沫在哪里？"他反问道。 软银的这些投资举措迫使公司在短期内 ...