作者 | 高允毅 自 8 月 GPT-5 发布以来，Codex展现出惊人的爆发力，用户增长 20 倍，每周处理数万亿 tokens，成为 了 Open AI 最受欢迎的编程智能体。 "Codex 能快速实现 20倍的增长，不只是因为模型变强了，还因为我们理解了，真正的智能体不是一个 模型，而是模型、API 和框架共同努力的结果。"在最新播客中，OpenAI 的编程智能体 Codex 产品负 责人 Alexander Embiricos 揭露背后的秘密。 比如，Codex 在长时任务能力上的突破。为了让它能够连续工作十几个小时甚至数天，团队设计了名 为"压缩"的机制——模型负责提炼关键信息，API 承接任务链路，框架负责稳定运行。三层像齿轮般咬 合，使 Codex 能够完成传统大模型难以支撑的长时编程任务。 正是这样的底层逻辑，让 Codex 在业务实战中有惊人表现。 Andrej Karpathy 曾公开分享，他被一个 bug 困住数小时，最终交给 Codex 处理，一小时内就完成了修 复。 Sora 团队更是依靠 Codex，在短短 28 天时间，从 0 到 1 完成 Android 应用的上线，直接冲到 ...

2024年10月29日（星期二），赛富时（Salesforce）旗下即时通讯平台Slack首席执行官丹妮丝·德雷瑟（Denise Dresser）出席在美国加利福尼亚州旧金山举办的TechCrunch Disrupt大会。TechCrunch Disrupt大会为期三天，聚焦科技 行业前沿资讯与发展动态，汇聚引领行业变革的思想领袖。图片来源：David Paul Morris/Bloomberg via Getty Images OpenAI上周二宣布，已选定Slack首席执行官丹妮丝·德雷瑟担任其首任营收主管。此举向持谨慎态度的 投资者传递了一个信息：这家ChatGPT的创造者正认真致力于从其AI技术中盈利。 奥尔特曼曾表示，ChatGPT目前拥有超过8亿周活跃用户。但这家估值5000亿美元的公司并未盈利，并 且已承诺向其AI系统所依赖的云计算提供商和芯片制造商支付超过1万亿美元的财务款项。 OpenAI可能无法赚取足够资金以满足甲骨文（Oracle）和英伟达（Nvidia）等支持者期望的风险，加剧 了投资者对AI泡沫的担忧。 OpenAI通过ChatGPT的高级订阅服务获得收入，但大多数用户使用的是免费版 ...

自 8 月 GPT-5 发布以来，Codex展现出惊人的爆发力，用户增长 20 倍，每周处理数万亿 tokens，成为 了 Open AI 最受欢迎的编程智能体。 "Codex 能快速实现 20倍的增长，不只是因为模型变强了，还因为我们理解了，真正的智能体不是一个 模型，而是模型、API 和框架共同努力的结果。"在最新播客中，OpenAI 的编程智能体 Codex 产品负 责人 Alexander Embiricos 揭露背后的秘密。 比如，Codex 在长时任务能力上的突破。为了让它能够连续工作十几个小时甚至数天，团队设计了名 为"压缩"的机制——模型负责提炼关键信息，API 承接任务链路，框架负责稳定运行。三层像齿轮般咬 合，使 Codex 能够完成传统大模型难以支撑的长时编程任务。 正是这样的底层逻辑，让 Codex 在业务实战中有惊人表现。 Andrej Karpathy 曾公开分享，他被一个 bug 困住数小时，最终交给 Codex 处理，一小时内就完成了修 复。 Sora 团队更是依靠 Codex，在短短 28 天时间，从 0 到 1 完成 Android 应用的上线，直接冲到 App Store ...

自 8 月 GPT-5 发布以来， Codex 展现出惊人的爆发力， 用户增长 20 倍 ，每周处理数万亿 tokens，成为了 Open AI 最受欢 迎的编程智能体。 "Codex 能快速实现 20 倍 的增长，不只是因为模型变强了，还因为我们理解了，真正的智能体不是一个模型，而是模型、 API 和框架共同努力的结果。" 在最新播客中，OpenAI 的编程智能体 Codex 产品负责人 Alexander Embiricos 揭露背后的秘 密。 比如，Codex 在长时任务能力上的突破。为了让它能够连续工作十几个小时甚至数天，团队设计了名为"压缩"的机制——模型 负责提炼关键信息，API 承接任务链路，框架负责稳定运行。三层像齿轮般咬合，使 Codex 能够完成传统大模型难以支撑的 长时编程任务。 正是这样的底层逻辑，让 Codex 在业务实战中有惊人表现。 作者 | 高允毅 Andrej Karpathy 曾公开分享，他被一个 bug 困住数小时，最终交给 Codex 处理，一小时内就完成了修复。 Sora 团队更是依靠 Codex，在短短 28 天时间，从 0 到 1 完成 Android 应用的上线 ...

Core Insights - Lucid Group has achieved seven consecutive quarterly delivery records, driven by demand for the Gravity model, which has positively impacted average selling prices [1][2][6] - The company has extended its liquidity runway into 2027, with pro forma liquidity reaching approximately $5.5 billion, allowing for continued growth and preparation for new model launches [8][6] - Profitability remains a significant challenge, with Q3'25 GAAP gross margin reported at approximately negative 99%, impacted by tariffs and inventory build [9] Delivery and Sales Performance - The demand for the Gravity model has led to a rise in average selling prices, with expectations for continued growth in deliveries into Q4 2025 [2][6] - Revenue projections indicate an increase from $1.22 billion in 2025 to $2.49 billion in 2026, reflecting a strong upward sales trajectory [3] Product Development and Efficiency - The Atlas powertrain is designed to enhance efficiency and reduce costs, with a focus on fewer parts and integrated systems, supporting the launch of a midsize model in late 2026 [5][7] - Management plans to leverage the Atlas powertrain to improve gross margins as production volumes increase [7] Financial Position and Liquidity - As of the end of Q3'25, the company's liquidity was approximately $4.2 billion, bolstered by a $2 billion undrawn term loan facility, extending the funding runway into the first half of 2027 [8] - This financial flexibility is crucial for scaling the Gravity model and preparing for the midsize launch while pursuing efficiency initiatives [8] Margin Challenges and Monitoring - Despite positive sales mix trends, the company faces margin pressures, with tariffs compressing reported margins by about 13 percentage points [9] - Investors are advised to monitor the availability of the Gravity Touring model and the impact of a second shift launched in October to support higher output [10]

SAN FRANCISCO (AP) — OpenAI said Tuesday it has picked Slack CEO Denise Dresser as its first chief of revenue, a message to wary investors that the ChatGPT maker is serious about making a profit from its artificial intelligence technology. OpenAI said Dresser will oversee global revenue strategy and “help more businesses put AI to work in their day-to-day operations.” Dresser had already spent more than a decade at Salesforce when the software pioneer announced in 2020 it was buying work-chatting service ...

Core Insights - Current AI browsers, such as OpenAI's Atlas and Perplexity's Comet, are not yet capable of replacing legacy browsers like Chrome, as they still exhibit bugs and struggle with straightforward tasks [1][22] - The goal of AI developers is to integrate chatbots into browsers and mobile operating systems, which could enhance ad targeting and create new revenue streams [1][22] - User behavior is shifting, with a growing need for tools that cater to both human and AI interactions, prompting developers to rethink their design strategies [2][22] User Behavior and Preferences - A survey indicated that 60% of users are only comfortable using generative AI for low-stakes tasks or topics they can verify, highlighting a cautious approach to AI adoption [4][22] - Users are increasingly combining traditional search methods with generative AI to ensure accuracy while saving time [5][22] - AI browsers are popular for summarizing content, such as long YouTube videos, and users are asking significantly more questions compared to traditional chatbots [6][22] Feature Requests and Expectations - There is a demand for advanced features like task scheduling, which would allow users to automate repetitive tasks [7][22][24] - Users express a desire for browsers to handle more complex tasks, such as filling out government forms and managing financial reports [8][24] Technical Challenges and Limitations - The current web infrastructure is primarily designed for human users, which hampers the performance of AI browsers in executing advanced features [13][14] - AI browsers often struggle with complex visual elements and can become stuck in loops or take excessive time to process tasks [15][16][17] - Legacy browsers have a head start in optimizing resource usage, while AI browsers tend to consume more computing power during advanced operations [17][22] Competitive Landscape - The AI browser market is evolving, with Google's Gemini 3 model outperforming OpenAI's systems, prompting a competitive response from AI developers [18][22] - Efforts to encourage online providers to create AI-specific versions of their services have faced challenges, with some companies hesitant to adopt [20][22] - Legal disputes have arisen as high-traffic firms seek to protect their content from being leveraged by AI companies, indicating a contentious environment [21][22] Trust and Adoption Barriers - Concerns about the trustworthiness of AI systems remain a significant barrier to wider adoption, particularly for tasks involving financial transactions [22][22] - Users are wary of potential conflicts of interest, such as whether an AI agent is providing the best deals or acting in its own interest [22][22]

Core Insights - MongoDB's stock is poised for significant growth following a strong fiscal Q3 2026 report, with shares increasing over 23% since the announcement [1] - Analysts from major firms have raised their price targets for MongoDB's stock, indicating positive market sentiment [3] Financial Performance and Growth Catalysts - In fiscal Q3 2026, MongoDB reported a 19% year-over-year revenue increase to $628.3 million, with non-GAAP net income per share improving to $1.32 from a loss of $1.16 in Q3 2024 [4] - The fully managed cloud database, Atlas, saw a revenue surge of 30%, now representing 75% of total revenue, driven by increased demand for AI-related workloads [6][7] - MongoDB's non-GAAP operating margin improved by 1 percentage point to 20%, and free cash flow increased by 300% year over year to $140 million, with a customer base growing to 62,500 [7] Updated Guidance - Management has raised fiscal 2026 revenue guidance to between $2.434 billion and $2.439 billion, up from $2.34 billion to $2.36 billion, and non-GAAP operating income guidance to $436.4 million to $440.4 million from $321 million to $331 million, reflecting strong demand trends and improved revenue visibility [8]

– NOT FOR DISSEMINATION IN THE UNITED STATES OR THROUGH U.S. NEWSWIRE SERVICES – Vancouver, BC & Boston, Massachusetts, Dec. 05, 2025 (GLOBE NEWSWIRE) -- Humanoid Global Holdings Corp. (“Humanoid Global” or the “Company”) (CSE:ROBO, FWB:0XM1, OTCQB:RBOHF), a publicly traded investment issuer focused on building and accelerating a portfolio of pioneering companies in the humanoid robotics and embodied AI sector, is pleased to announce that Marc Theermann, chief strategy officer of Boston Dynamics, has joined ...

Core Insights - Humanoid Global Holdings Corp. has appointed Marc Theermann, Chief Strategy Officer of Boston Dynamics, to its Technical Advisory Committee, enhancing its expertise in humanoid robotics and embodied AI [1][3][4] Company Overview - Humanoid Global is a publicly traded investment issuer focused on building a portfolio of pioneering companies in the humanoid robotics and embodied AI sector, providing liquidity and access to an actively managed portfolio [5] - The company takes a long-term, partnership-oriented approach, offering capital and strategic consultation on go-to-market strategies, regulatory pathways, and transaction advisory [5] Appointment of Marc Theermann - Marc Theermann will provide strategic guidance on emerging developments in humanoid robotics, including manipulation, perception, mobility, and human-robot interaction [3] - His role includes identifying research partners, academic collaborators, and technical talent, as well as evaluating potential investment opportunities in the humanoid robotics landscape [3][4] - Theermann's experience in commercialization and platform development is expected to be valuable for Humanoid Global as it evaluates opportunities in the sector [4] Compensation and Stock Options - Humanoid Global has granted Marc Theermann 50,000 incentive stock options at an exercise price of $1.29, vesting over three years, and 60,000 restricted share units (RSUs), with 10,000 RSUs vesting immediately [4] - The remaining RSUs will vest in a similar manner as the stock options, with a monthly fee of $7,500 to be paid to Theermann [4]