Meta首席人工智能科学家、图灵奖得主Yann LeCun正为其新创立的AI公司寻求5亿欧元融资，此举将使 这家人工智能公司在正式推出之前估值达到约30亿欧元。 据英国金融时报援引知情人士透露，即将于年底离开Meta的LeCun已任命法国健康科技初创公司Nabla 创始人Alexandre LeBrun担任新公司首席执行官。该企业命名为"先进机器智能实验室"（Advanced Machine Intelligence Labs），计划于明年1月公布详细信息，LeCun将出任执行主席。 这是今年人工智能领域又一笔备受关注的高额融资案例。此前，OpenAI联合创始人Ilya Sutskever于4月 为其成立仅一年、尚未推出产品的AI公司Safe Superintelligence成功筹集20亿美元，估值高达320亿美 元。 瞄准超级智能AI系统 Nabla联合创始人出任CEO Nabla联合创始人Delphine Groll在一份声明中表示，经董事会批准的过渡计划中，公司联合创始人兼首 席执行官Alex LeBrun将卸任现职，并出任AMI Labs的首席执行官。 与此同时，Nabla已与AMI Labs建立战 ...

Core Viewpoint - Despite not raising the full-year guidance for fiscal year 2026 as some investors expected, Goldman Sachs reaffirmed a "Buy" rating for Broadcom, citing the company's strengthening dominance in the custom chip sector and robust fundamentals in its AI business [1][2]. Financial Performance - Broadcom reported a strong fourth-quarter revenue of $18 billion, exceeding market expectations of $17.5 billion. The guidance for the first quarter of fiscal year 2026 is set at $19.1 billion, significantly above the analyst forecast of $18.3 billion. This growth is primarily driven by a 74% year-over-year increase in AI semiconductor revenue [1][5]. - The AI semiconductor revenue reached $6.5 billion, surpassing the expected $6.2 billion, while total semiconductor solutions revenue was $11.1 billion, above the anticipated $10.7 billion. Infrastructure software revenue was $6.9 billion, slightly exceeding expectations [5]. Market Reaction and Guidance - The market's reaction to the earnings report may be mixed, as there is disappointment over the lack of an updated full-year guidance for fiscal year 2026. Analysts noted that the absence of an upward revision could lead to short-term stock price pressure [1][3]. - Goldman Sachs' analysts predict that Broadcom's AI revenue growth for fiscal year 2026 will actually exceed 100%, despite management's conservative stance not reflecting the actual business momentum [3]. Customer Expansion and Orders - Broadcom has made significant progress in customer expansion, maintaining strong momentum with its largest client, Google, and announcing a new major customer. The company has secured a fifth XPU customer and received an additional $11 billion order from Anthropic for fiscal year 2026 [4]. - The backlog of AI orders has reached $73 billion for the next 18 months, indicating strong demand, with additional orders expected to increase this figure [4]. Profitability Trends - Broadcom's gross margin for the fourth quarter was 77.9%, slightly above market expectations. The adjusted EBITDA margin guidance for the first quarter of fiscal year 2026 is set at 67% [5]. - Goldman Sachs noted that as Broadcom begins delivering full-rack solutions to Anthropic and potentially OpenAI in the second half of fiscal year 2026, there may be some dilution in gross and operating margins due to a higher proportion of direct components in these solutions [5]. Valuation Logic - Goldman Sachs raised Broadcom's 12-month price target from $435 to $450, based on an increase in AI revenue expectations and improved visibility into industry cycles. The estimated EPS was adjusted from $11.50 to $12.00, maintaining a 38x P/E multiple [2][7].

Dec 9 (Reuters) - Microsoft said on Tuesday it was investing more than C$7.5 billion ($5.42 billion) in Canada over the next two years, its ​latest in a series of investments in artificial intelligence infrastructure globally. New ‌capacity under the investment will begin to come online in the second half of 2026, Microsoft ‌said, adding that its total estimated investment in Canada amounts to C$19 billion between 2023 and 2027. Microsoft has pledged hefty investments worldwide this year, as the company ...

硅谷风险投资公司a16z近日发布了名为 《State of AI:An Empirical 100 Trillion Token Study》的重量级年度报告。 这分析了AI模型路由平台OpenRouter上超过100万亿个真实生产环境中的Token使用数据，系统性地揭示了大语言模型（LLM）的实际使用图 景。 这是迄今为止规模最大、最全面的AI使用实证分析，数据覆盖全球60余家模型提供商的300多个模型，展现了AI技术从实验室走向真实世界的 完整轨迹。 研究发现，AI领域正在经历三大根本性转变：从单一模型竞争走向多元化生态系统； 从简单文本生成迈向智能体推理范式；从西方中心向全球分布式创新格局演进。 整个报告中，还有一些有趣的发现： DeepSeek模型展现出的"回旋镖效应"。部分用户在尝试其他模型后，会重新回归DeepSeek。 中国开源力量崛起：从2024年底几乎可以忽略不计的市场份额（周使用量占比低至1.2%），到2025年后期在某些周度达到近30%的占比； 超过一半的开源模型使用量流向了角色扮演、故事创作等创意对话场景； 推理模型所处理的Token量已占总量的50%； 编程相关的查询量在2025年实 ...

【文/观察者网 齐倩】 作为加拿大科技初创公司Cohere首席执行官，艾丹·戈麦斯自满于能与美国合作。他还断言美加联手定 能胜过中国，给出的理由竟是所谓"民主国家"不愿用中国技术。 艾丹·戈麦 斯近期接受美媒采访 视频截图 报道提到，随着AI发展迅速，科技投资者已经在该领域投资数百亿美元。但近来，这些投资者开始要 求微软和谷歌母公司Alphabet等科技公司展现更好回报。对此，戈麦斯辩解说，过去几年，随着大规模 AI应用扩展，"我们看到模型改进速度正在放缓"。 据路透社消息，当地时间12月4日，戈麦斯在纽约参加路透NEXT大会时做出的上述表述。 据介绍，Cohere总部位于多伦多，专注于构建企业专属人工智能（AI）模型。 戈麦斯在会上称，在与全球采用AI的经济体合作方面，美国和加拿大处于"难以置信的有利位置"，而这 使得美加在全球AI竞赛中"领先于中国"。 戈麦斯承认，中国已推出性能极高的AI模型，缩小了与一些顶尖闭源大语言模型之间的差距。但他随 后声称："真正重要的是，谁是这项技术的主要服务提供者——不是谁先拿到技术，而是谁能大规模商 业化。美国和加拿大处于绝佳位置，成为全球采用这项技术的伙伴。" 作为技术 ...

Core Viewpoint - The CEO of Canadian startup Cohere, Aidan Gomez, asserts that the collaboration between the U.S. and Canada will surpass China in the AI sector, emphasizing that democratic nations are reluctant to rely on Chinese technology [1][3]. Group 1: Company Insights - Cohere, based in Toronto, focuses on building enterprise-specific AI models and is positioned favorably in the global AI competition [1]. - Gomez claims that the U.S. and Canada are in an "incredibly advantageous position" for global AI adoption, despite acknowledging that China has developed high-performance AI models [1][3]. - The company argues that the key factor is not who develops the technology first, but who can commercialize it on a large scale [1]. Group 2: Industry Trends - The rapid development of AI has led to significant investments, with tech investors pouring hundreds of billions into the sector [3]. - There is a growing demand from investors for better returns from major tech companies like Microsoft and Alphabet, indicating a shift in expectations [3]. - Concerns about the risks associated with AI technology are rising, but Gomez downplays apocalyptic narratives, suggesting that society is adapting to the realities of AI [3]. Group 3: Competitive Landscape - The AI competition between the U.S. and China is intensifying, with Chinese startups like DeepSeek gaining traction and major companies like Alibaba and Baidu accelerating their AI product launches [3][4]. - U.S. tech giants are investing heavily in enhancing their computing capabilities and AI infrastructure to maintain leadership in the sector [4]. - Nvidia's CEO Jensen Huang has expressed concerns about China's potential to win the AI race, highlighting the competitive pressures faced by U.S. companies [4].

【文/观察者网 齐倩】 作为加拿大科技初创公司Cohere首席执行官，艾丹·戈麦斯自满于能与美国合作。他还断言美加联手定 能胜过中国，给出的理由竟是所谓"民主国家"不愿用中国技术。 据路透社消息，当地时间12月4日，戈麦斯在纽约参加路透NEXT大会时做出的上述表述。 据介绍，Cohere总部位于多伦多，专注于构建企业专属人工智能（AI）模型。 戈麦斯在会上称，在与全球采用AI的经济体合作方面，美国和加拿大处于"难以置信的有利位置"，而这 使得美加在全球AI竞赛中"领先于中国"。 报道提到，随着AI发展迅速，科技投资者已经在该领域投资数百亿美元。但近来，这些投资者开始要 求微软和谷歌母公司Alphabet等科技公司展现更好回报。对此，戈麦斯辩解说，过去几年，随着大规模 AI应用扩展，"我们看到模型改进速度正在放缓"。 对于社会对AI技术风险的担忧加剧，戈麦斯回应表示："我个人并不太相信那些'终结者'、末日之类的 科幻叙事，这些说法后来变得不受欢迎，因为人们已经面对了AI技术的现实。" 路透社称，进入2025年后，中美两国间AI竞赛升温。今年1月，中国AI初创公司DeepSeek迅速走红，此 后，包括阿里巴巴和百 ...

ROI of AI - MIT研究表明，95%的企业在实施AI时未获得投资回报率，而Wharton的研究则表明74%的企业获得了投资回报率，行业对AI的投资回报率存在争议 [2] - 行业普遍认为AI的未来是光明的，但目前仍处于实验阶段，以找到实际应用并扩大规模的方法 [3] - Wharton的研究显示，74%的企业声称从AI中获得投资回报率，但副总裁及以上级别的人比经理及以下级别的人更倾向于认为他们获得了投资回报率 [8] - 衡量投资回报率的方式多种多样，包括减少人员需求、创造竞争优势以加速收入增长，以及提高效率以降低毛利率 [11] - Wharton的研究显示，年收入超过20亿美元的大公司从AI中获得投资回报率的比例为57%，低于较小公司 [25] Compute Cost and Infrastructure - Deep Seek的出现将推理模型的成本降低了约40倍 [14] - Grock致力于降低计算成本，其芯片在运行AI推理时，不仅性能领先市场，而且成本也具有差异化 [14][16] - 行业普遍认为，目前行业面临供应限制，需要部署更多容量以满足所有AI用例 [17] - 到2028年，美国四分之一的能源将用于运行AI工作负载 [19] AI Strategy and Implementation - 企业需要制定战略，明确在公司内部署AI的领域，并考虑如何衡量投资回报率 [22] - 企业应建立一个框架，围绕他们想要做什么、如何做、何时做，以及投资多少资金，并在后端进行衡量 [23] - 较小的公司更容易适应，能够更快地实施计划并在公司内部的多个不同职能部门中实施，从而获得更高的投资回报率 [26] - AI解决方案可以立即产生价值，但需要重新配置流程，而不是事后设计 [30] AI Bubble and Future Growth - 行业专家认为，对AI基础设施的投资不会后悔，但需要定义“泡沫”的含义，包括基础设施建设和公司估值 [32][33] - 行业专家预测，未来会出现一批万亿美元级别的公司，风险投资者正在选择他们的赌注 [34] - 行业专家认为，AI的某些领域可能存在泡沫，但不会影响整个行业，即使出现调整，也不会影响每家公司 [36] - Anthropic的CEO表示，该公司几年前的收入为10亿美元，预计到今年年底的运行率将接近100亿美元 [40]

Core Insights - SAP has launched the EU AI Cloud, consolidating its digital sovereignty initiatives into a framework tailored for Europe, offering customers various options for data sovereignty and deployment [1][3] - The platform is designed to meet EU data residency and regulatory compliance requirements, allowing enterprises to incorporate AI into core business processes while adhering to local regulations [2][3] Deployment Options - The EU AI Cloud provides multiple deployment options, including SAP Sovereign Cloud, on-site models, and partnerships with approved European cloud providers, ensuring compliance with European regulations [3][5] - AI workloads are hosted in European data centers, ensuring independence from non-European hyperscalers, which is crucial for maintaining data sovereignty [4] Partnerships and Collaborations - SAP has partnered with Cohere to integrate agentic AI capabilities into the SAP Business Technology Platform, enhancing AI integration for enterprises with data residency constraints [2] - Tata Consultancy Services (TCS) has been selected to support SAP's cloud and generative AI transformation under a five-year agreement, focusing on streamlining IT operations and enhancing AI capabilities [6][7]

Core Insights - SAP SE has launched the EU AI Cloud, a sovereign AI and cloud offering designed specifically for Europe, which allows customers to choose their level of sovereignty and deployment options [1][2][4] - The collaboration with Cohere aims to enhance AI capabilities for European enterprises, integrating advanced AI models into the SAP Business Technology Platform (SAP BTP) [2][3] - The EU AI Cloud is supported by a robust ecosystem of partners, enabling flexible deployment options and compliance with European data protection regulations [3][5] Group 1: EU AI Cloud Overview - The EU AI Cloud provides a full-stack sovereign cloud offering, ensuring compliance with EU data residency and operational requirements [1][5] - Customers can deploy AI models on SAP's infrastructure or trusted European partners, maintaining control over their data and compliance [5][8] Group 2: Partnership and Ecosystem - SAP's partnership with Cohere will unlock agentic AI capabilities, allowing enterprises to integrate AI into their core business processes without compromising on sovereignty [2][4] - The ecosystem includes leading partners like Mistral AI and OpenAI, facilitating the development and scaling of AI-powered applications [3][4] Group 3: Deployment Flexibility - EU AI Cloud offers various deployment options, including SAP Sovereign Cloud and on-site solutions, tailored to meet specific regulatory and operational needs [5][8] - The infrastructure is designed to ensure that all data remains within the EU, aligning with European data protection standards [8]