Anthropic
Search documents
Amazon Bets $50 Billion On OpenAI In Landmark AI Partnership - Amazon.com (NASDAQ:AMZN)
Benzinga· 2026-02-27 14:13
Amazon.com Inc. (NASDAQ:AMZN) and OpenAI announced a sweeping multi-year strategic partnership Friday morning, with Amazon committing to invest $50 billion in the ChatGPT maker, starting with an initial $15 billion, followed by another $35 billion when certain conditions are met. A New Layer For Cloud AIAt the heart of the deal is a jointly developed Stateful Runtime Environment powered by OpenAI’s models, set to be available through Amazon Bedrock. Unlike traditional cloud AI integrations, stateful environ ...
OpenAI完成1100亿美元融资,获亚马逊、英伟达、软银投资
Xin Lang Cai Jing· 2026-02-27 13:55
要点速览 OpenAI表示,其最新一轮融资已筹集1100亿美元。 OpenAI周五宣布, 亚马逊 投资500亿美元,英伟达投资300亿美元, 软银 投资300亿美元。 OpenAI持续快速增长,但这家AI巨头正面临来自Anthropic和 谷歌 日益激烈的竞争。 尽管OpenAI继续在消费级AI市场保持领先,但也面临着来自谷歌Gemini日益激烈的竞争,并正努力加 强其面向企业市场的产品供应——其竞争对手Anthropic在该领域已取得先发优势。 消息人士称,OpenAI预计其2030年总收入将超过2800亿美元,其中消费者和企业业务贡献几乎持平。 由于信息未公开,消息人士要求匿名。 OpenAI的这轮融资是历史上规模最大的私募融资,也创下了晚期科技公司估值的新高。OpenAI去年首 次以由软银领投的400亿美元融资打破了纪录。其竞争对手Anthropic以300亿美元的最近一轮融资紧随 其后,而xAI最近一轮融资额为200亿美元。 OpenAI已完成一轮1100亿美元的融资,其规模是一年前创下私营科技公司纪录的上轮融资的两倍多。 OpenAI在周五的一份新闻稿中表示,亚马逊投资500亿美元,英伟达投资300 ...
Anthropic称不会在AI用于国防的问题上退让
Xin Lang Cai Jing· 2026-02-27 13:38
Anthropic不会在AI用于国防的问题上退让。CEO达里奥·阿莫代昨日表示,这家AI初创公司"不能凭良 心"允许国防部在所有合法用例中不受限制地使用其模型,而五角大楼希望如此。 Anthropic一直要求国防部承诺,其模型不会被用于全自动武器或对美国民众进行大规模国内监视—— 五角大楼不愿同意这些限制。 Anthropic昨日表示,谈判仍在进行中。国防部长皮特·赫格塞斯将今日设定为Anthropic同意其部门要求 的最后期限,并威胁称,如果不同意,将把该公司列为"供应链风险"或援引《国防生产法》。 责任编辑:张俊 SF065 Anthropic不会在AI用于国防的问题上退让。CEO达里奥·阿莫代昨日表示,这家AI初创公司"不能凭良 心"允许国防部在所有合法用例中不受限制地使用其模型,而五角大楼希望如此。 Anthropic一直要求国防部承诺,其模型不会被用于全自动武器或对美国民众进行大规模国内监视—— 五角大楼不愿同意这些限制。 Anthropic昨日表示,谈判仍在进行中。国防部长皮特·赫格塞斯将今日设定为Anthropic同意其部门要求 的最后期限,并威胁称,如果不同意,将把该公司列为"供应链风险"或援 ...
Google, OpenAI Staff Stand With Anthropic, Demand 'Red Lines' In Pentagon AI Contracts: 'We Will Not Be Divided' - Alphabet (NASDAQ:GOOG), Alphabet (NASDAQ:GOOGL)
Benzinga· 2026-02-27 13:19
Alphabet’s Google (NASDAQ:GOOGL) (NASDAQ:GOOG) and OpenAI employees are raising their voices against the company’s plans to collaborate with the Pentagon on artificial intelligence technology.Over 100 Google staff members penned a letter to management expressing their concerns. They called for the establishment of clear boundaries or “red lines” in government contracts, akin to those pursued by Anthropic, reported the New York Times on Thursday.The employees took issue with the possible use of Google’s Gemi ...
Netflix drops its WBD bid, Block layoffs, Anthropic's DOD deadline and more in Morning Squawk
CNBC· 2026-02-27 13:08
Group 1: Paramount Skydance and Warner Bros. Discovery - Paramount Skydance is making a hostile takeover bid for Warner Bros. Discovery valued at $108.4 billion, which has led Netflix to withdraw from its deal to acquire some of WBD's assets after WBD's board deemed Paramount's offer superior [2][6] - Paramount's new all-cash bid is set at $31 per share, surpassing Netflix's current agreement with WBD, prompting Netflix to state that matching the offer is no longer financially attractive [6] Group 2: Employment Changes in Tech - Block, the fintech company, announced layoffs of over 4,000 employees, approximately half of its workforce, to position itself for long-term growth, resulting in a 20% surge in its shares during extended trading [3][4] - eBay also announced a reduction of about 800 roles, or 6% of its staff, as part of its investments in AI, indicating a trend of staffing changes across the tech industry [4] Group 3: Women's Wealth Growth - Women's cumulative investible assets in the U.S. are expected to nearly double from 2023 to 2030, with a significant transfer of over $100 trillion in wealth anticipated through 2048, marking a substantial shift in wealth control [10]
X @Decrypt
Decrypt· 2026-02-27 12:51
Claude Creator Anthropic Says It Won’t Lift AI Safeguards Despite Threats From the Pentagon!Anthropic CEO’s said the company will not comply with Defense Department demands as the Pentagon weighs whether to label the company a “supply chain risk.” https://t.co/TascgXJmgW ...
美股前瞻 | 三大股指期货齐跌,PPI数据今晚来袭
智通财经网· 2026-02-27 12:47
1. 2月27日(周五)美股盘前,美股三大股指期货齐跌。截至发稿,道指期货跌0.65%,标普500指数期货跌0.48%,纳指期货跌0.48%。 | ■ US 30 | 49,175.30 | 49,364.60 | 49,108.70 | -324.20 | -0.65% | | --- | --- | --- | --- | --- | --- | | ■ US 500 | 6,875.90 | 6,905.00 | 6,868.50 | -33.00 | -0.48% | | ■ US Tech 100 | 24,913.00 | 25,051.30 | 24,875.30 | -121.40 | -0.48% | 2. 截至发稿,德国DAX指数涨0.23%,英国富时100指数涨0.49%,法国CAC40指数跌0.16%,欧洲斯托克50指数跌0.05%。 | 1 德国DAX30 | 25,334.95 | 25,407.75 | 25,231.95 | +57.25 | +0.23% | | --- | --- | --- | --- | --- | --- | | 器 英国富时100 | 10,900. ...
'COMMON-SENSE': GOP rep TORCHES critics, doubles down on SAVE Act
Youtube· 2026-02-27 12:31
Group 1: Department of Homeland Security Shutdown - The Department of Homeland Security (DHS) is entering its third week of shutdown, affecting over 57,000 employees, with some classified as "accepted" not receiving pay until the shutdown ends [1][3] - The shutdown is described as a partisan issue, with calls for reforms in agencies like TSA, FEMA, and Border Patrol without interrupting their pay [4][5] Group 2: Save America Act - The Save America Act, which requires proof of citizenship to vote, is stalled in the Senate, with bipartisan support indicated by 80% of Americans finding it sensible [2][19] - Concerns are raised about potential voter suppression, as approximately 40% of Americans may not have the required documents like birth certificates or passports [16][19] Group 3: AI and National Security - Anthropic's CEO has refused to allow their AI products to be used for mass domestic surveillance, raising questions about the implications of AI in military and surveillance contexts [6][8][13] - The Pentagon is urged to maintain a competitive edge in AI, especially against advancements made by China in military applications [7][11]
政协委员周鸿祎:AGI正稳步实现,智能体重塑网络生态
2 1 Shi Ji Jing Ji Bao Dao· 2026-02-27 12:25
Core Insights - The commercialization of general artificial intelligence (AGI) is becoming clearer by 2026, with a focus on building intelligent agent ecosystems and enhancing reasoning capabilities [1][3] - AI is entering the cybersecurity market, reshaping the attack and defense systems, indicating a significant trend in the industry [1][3] Group 1: AGI Development - AGI is being redefined, with current AI capabilities surpassing the average human skill level, rather than requiring a "super genius" [3] - The Seedance video generation model exemplifies AGI capabilities, demonstrating significant potential in the entertainment industry [3] - Effective use of AI involves creating specialized intelligent agents that can engage in deep reasoning through role-playing and collaborative debate [3] Group 2: Intelligent Agents in Internet Economy - The rise of intelligent agents is leading to the emergence of an "agent economy," where agents will facilitate automatic price comparisons and transactions on e-commerce platforms [5] - This new business model raises questions about identity verification and accountability, particularly regarding errors made by deployed agents [5] - Intelligent agents are expected to fundamentally alter existing internet products and business models, potentially leading to parallel systems for human interaction and API access [5] Group 3: Cybersecurity Transformation - AI tools like Claude code security are revolutionizing the cybersecurity industry by efficiently scanning for vulnerabilities and generating patches, causing stock declines for traditional security firms [7] - The efficiency of AI in programming may lead to an overwhelming amount of code that humans cannot effectively manage, necessitating specialized AI tools for security [7][8] - The traditional cybersecurity model, which focuses on post-attack defense, is being challenged as AI can potentially eliminate many vulnerabilities during the coding phase [8] Group 4: Future of Cyber Attacks - Future cyber attacks are expected to evolve into "hacker agents," which will automate and scale attack methods beyond human capabilities [9] - The traditional defense strategies will likely collapse under the pressure of automated hacker agents, necessitating a shift in cybersecurity approaches [9] - Companies like 360 Group are adopting intelligent agents to enhance security operations, including vulnerability detection and automated penetration testing [9]
按参数算,我们1300克的人脑相当于多大的AI模型?
3 6 Ke· 2026-02-27 12:25
01 按参数算,人脑相当于多大的模型? 答案是:要看怎么算。 如果只看神经元的个数,人脑大概是860亿个神经元,也就是86B的模型,并不大。 参考一下,DeepSeek V3是671B,Kimi K2.5大概1000B,即1T; 但事实上人脑每个神经元又有7000个突触,从技术的角度类比,颗粒度更小的突触才更像AI模型的权重参数。 如果这么算,860亿*7000,那么人类大脑相当于大约600T模型。 而这么大的模型,今天的硬件肯定暂时还跑不动。 这么类比略糙。 但也说明——咱们这颗脑子的架构还是很复杂的,属于先进制程。 有点牛逼。 02 那么,大脑的制程到底有多先进呢? 我随即问了Claude opus 4.6和Gemini 3.1 Pro一个问题(实在受不了GPT无比谄媚的风格)—— "如果人脑是一块芯片,那么它的制程是几纳米的?" 他们的答案出奇一致: 如果看神经元细胞体直径的直径,大概 10000-100000 纳米 (10-100微米)。 这么看大脑相当于几十年前的电子管计算机。 这TM也太落后了。 但逻辑显然不是这样的: 神经元并非一个简单的开关,它更像处理器的一个核,真正的开关和信号传递发生在突触 ...