蚂蚁阿福与OpenAI health打开C端AI医疗想象空间。 硬·AI 作者 | 申思琦 编辑 | 硬 AI 医药投资圈曾流行一种偏见，投资人普遍认为数字医疗是个伪命题。 他们眼中的医疗需求存在一个"不可能三角"：低频、高门槛、非标品。普通人不会天天看病，医生需要十年培养，每个人的病历都独一无二。 这导致互联网医疗平台常年陷于买流量的泥潭——获客成本极高，用户留存极低。 然而，这一刻板印象正在被AI时代的数据洪流无情击碎，C端的AI医疗需求实际上是极其惊人的，它一直都存在，只是过去缺乏一个足够低门槛、低成本且足够 智能的交互容器来承接。 当交互成本降至零，且反馈质量达到准专业级时，AI医疗这种沉默的刚需瞬间爆发了。 在两个AI超级巨头的动作中看到了这种爆发，一个是中国的蚂蚁集团，一个是美国的OpenAI。 在中国， 蚂蚁集团旗下的"阿福"，其 月活跃用户数已在一个月内翻倍突破3000万，单日提问量超过1000万次 。 大洋彼岸，OpenAI于2026年1月7日正式推出OpenAI Health。 OpenAI的数据显示， 全球每周有超过2.3亿人次在ChatGPT上咨询健康问题。这甚至发生在该产品推出之前， ...

关键字：韧 韧性闯关奔赴星海 投资于人赢得未来 在中央关于制定"十五五"规划的《建议》中，确立了"十五五"我国"坚持扩大内需这个战略基点，坚持 惠民生和促消费、投资于物和投资于人紧密结合"的指导思想。这意味着我国未来发展将转向以促进人 的全面发展为目的的内需主导模式。 "投资于人"是指将更多的财政资金和公共资源投向教育、就业、医疗、文化、社会保障等民生领域，投 入到人的知识获取、能力提升、健康维护、职业发展和潜力开发中，将"人力资源"转化为可持续增值 的"人力资本"，将"人口红利"转化为"人才红利"。"投资于人"所获得的劳动者个体素质能力提升和人力 资本价值增值，不仅是人的全面发展的重要内容，也是社会财富增值乃至经济社会发展的重要源 泉。"投资于人"通过为广大劳动者提供更加丰富的公共产品供给和社会保障，降低生活成本，进而增强 全社会的心理预期，提高全社会的消费信心，以消费潜力释放和人力资本提升推动经济高质量发展。同 时，"投资于人"所作用的教育、医疗、文化、社会保障等领域也是社会主义现代化建设的重要领域，是 联合国提出的衡量一国现代化发展水平的人类发展指数（HDI）的重要指标。可以说，"投资于人"投的 是当下 ...

编辑｜张倩、陈陈 当智能体（Agent）开始深度介入人类世界，关于豆包 AI 手机的讨论可能只是个开始。 在此之前，手机、电脑软件都是给人用的 —— 人负责一步步操作，系统负责把信息存好、算好。但现在，Agent 开始接过这些操作：你只需要说清楚目标，它就 能自己去打开应用、填信息、做选择，最后把结果交给你确认。 这就引发了一个问题：当人不再需要亲自点每一步，原本围绕「人来操作」设计的软件、系统还有没有存在的必要？除了豆包 AI 手机这样的 to C 场景，其实企业 也在争论这个问题。 最近的讨论集中在一个叫「记录系统（Systems of record）」的东西上面。有人说 Agent 杀死了记录系统，也有人说 Agent 只是提高了「好的记录系统」的标准， 还有人说，围绕 Agent 执行流程而搭建的新型「记录结构」，背后隐藏着万亿美元的机会。 那么，记录系统到底是什么？围绕它的机会存在于哪里？我们总结了几篇相关文章，试图详细分析这些问题。 记录系统已死？ 企业里的记录系统，说白了就是公司的「总账本」和「黑匣子」。谁做了什么、什么时候做的、数据改过几次、流程走到哪一步，都会被它原样记下来，方便之 后对账、 ...

wo[Table_Title] Research Report 21 Dec 2025 中国电子 China (Overseas) Technology 火山引擎 FORCE 大会追踪（2）： Agent 规模化落地，方舟与企业底座升 级 Volcengine FORCE(2): Scaling Agents to Production: Upgrades to Volcano Ark and Enterprise Foundations 姚书桥 Barney Yao 吕小潼 Xiaotong Lyu barney.sq.yao@htisec.com xt.lyu@htisec.com [Table_yemei1] Flash Analysis [Table_summary] (Please see APPENDIX 1 for English summary) 风险: 事件 2025 年 12 月 18 日，火山引擎在 2025 冬季 FORCE 原动力大会上发布了面向多模态 Agent 场景优化的豆包大模型在 2025 FORCE 原动力大会开发者日（上海）上，火山引擎围绕智能体（Agent）规模化生产与 ...

Core Insights - OpenAI researcher Yao Shunyu has joined Tencent, igniting discussions in the AI community [1] - Tencent has upgraded its large model research framework, establishing new departments to enhance its capabilities [2][3] Group 1: Organizational Changes - Tencent has formed the AI Infra Department and AI Data Department to strengthen its large model research and core capabilities [2] - Yao Shunyu has been appointed as the Chief AI Scientist, reporting to Tencent's President Liu Chiping, and will also lead the AI Infra Department and the large language model department [2][5] Group 2: Department Responsibilities - The AI Infra Department will focus on building technical capabilities for large model training and inference platforms, emphasizing distributed training and high-performance inference services [3] - The AI Data Department and Data Computing Platform Department will be responsible for constructing data and evaluation systems for large models and integrating big data with machine learning [4] Group 3: Yao Shunyu's Background - Yao Shunyu is a prominent young researcher in the field of artificial intelligence, particularly in the area of intelligent agents [6] - Prior to joining OpenAI, he made significant contributions in the field of language intelligent agents and has a total citation count exceeding 19,000 for his papers [7]

然而，下一波真正的浪潮已经涌现：智能体（Agent）时代。这里的"智能体"，不是简单的聊天机器 人。它指的是 以大型语言模型为认知引擎，具备自主决策、目标导向和环境交互能力的 AI 系统。 从 ChatGPT 到 Claude，从文心一言到 DeepSeek，我们已习惯与各种 Copilot 协作。它们能力惊 人，但本质仍是 被动响应 的工具——我们发出指令，它们返回结果。 "我看到很多开发者卡在'只会调 API'的层面，这非常可惜。AI 智能体的底层，是一套精密的 系统工程。掌握它，意味着你能从'工具的使用者'转变为'智能系统的创造者'。这不仅是技能 的提升，更是思维模式的跃迁。" 想象一下： 这就是 Agentic AI ——未来最炙手可热的技术方向，也是拉开下一代 AI 应用差距的关键。 热潮之下，开发者真正的 挑战是什么？ 智能体开发就像一座冰山。水面之上，是大家熟悉的"调 API、写 Prompt"；水面之下，才是真正的 复杂性所在： 这些，才是考验开发者能否将炫酷概念转化为真实价值的核心能力。面对这片充满机遇但略显复杂的 海域，一位资深的技术"引路人"—— 王延飞老师 ，分享了他的观察。 作为一位 ...

今年8月，国务院印发《关于深入实施"人工智能+"行动的意见》为我国推动人工智能与经济社会各行业 各领域广泛深度融合提供了指引。10月28日发布的《中共中央关于制定国民经济和社会发展第十五个五 年规划的建议》中再次明确："深入推进数字中国建设""加快人工智能等数智技术创新""全面实施'人工 智能+'行动，以人工智能引领科研范式变革，加强人工智能同产业发展、文化建设、民生保障、社会治 理相结合，抢占人工智能产业应用制高点，全方位赋能千行百业"。 从线上人工智能（以下简称"AI"）大模型与教育、医疗、金融等各行各业深度绑定，持续刷新行业效率 上限；到线下具身智能机器人在工厂协作、社区养老、家庭服务中崭露头角，为人类生产生活带来无限 可能……2025年，人工智能正以前所未有的速度穿透虚拟与现实、串联技术与产业，也让人们不禁畅 想，AI技术的未来发展还能带来怎样的惊喜。 不过，在AI技术打破人机边界的背后，安全治理也成为不可忽视的命题。姚期智提醒，AI算法潜在的 不可靠性可能引发隐私泄露、冲击社会价值伦理等风险。"目前中外正在探索将AI与密码学、博弈学等 理论结合的交叉领域，凝聚国际共识，携手构建AI治理协议。"姚期智 ...

Core Insights - AWS is at a pivotal moment, focusing on transforming AI from an assistant role to a more capable agent role, aiming to deliver real business value to enterprise customers [1][18]. Group 1: Computing Power - AWS has adopted a more pragmatic and aggressive strategy in computing power, significantly reducing costs with self-developed chips and breaking physical boundaries to accommodate large clients who prefer not to migrate to the cloud [4]. - The introduction of Trainium 3 UltraServers has improved inference efficiency by five times compared to its predecessor, with Trainium 4 promising a further sixfold performance increase [4][27]. - AWS AI Factories have been launched to address data sovereignty concerns by allowing clients to build AWS's computing infrastructure directly in their data centers [4]. Group 2: Model Development - AWS has completed its Amazon Nova self-developed model family with the release of the Amazon Nova 2 series, which includes the first multimodal model capable of processing text, images, audio, and video inputs [6]. - The Amazon Nova Forge introduces the concept of "open training models," allowing enterprises to inject proprietary data during the final stages of model pre-training, enhancing the model's capabilities without losing core competencies [6][37]. Group 3: Application Layer - AWS is addressing the uncontrollable nature of AI agents by implementing a robust policy framework, AgentCore Policy, to ensure agents act as reliable productivity tools [7][45]. - The Frontier Agents series has been introduced, which includes autonomous agents capable of performing tasks such as bug fixing and security scanning, indicating a shift in software engineering lifecycle management [7][41]. - The AgentCore platform is designed to facilitate the secure and scalable deployment of agents, with features that allow for real-time monitoring and control of agent actions [41][44]. Group 4: Business Growth and Infrastructure - AWS reported an annual revenue of $132 billion, with a growth rate of 20%, indicating strong business performance and market leadership in cloud computing [10]. - The company has expanded its global data center network, adding 3.8 GW of capacity in the past year, which is the largest in the industry [13]. - AWS's collaboration with various partners, including startups and established companies, highlights its role in driving innovation across multiple sectors [14].

Core Viewpoint - ByteDance is developing the Doubao mobile assistant, which integrates AI capabilities to automate complex mobile tasks, enhancing user experience and interaction with smartphones [3][36]. Group 1: Doubao Mobile Assistant Features - The Doubao mobile assistant allows users to perform tasks that typically require multiple screen taps through voice commands or minimal manual input [3][5]. - It can execute cross-application operations, such as comparing prices across different food delivery platforms and automatically gathering information [5][9]. - The assistant has a "correction mechanism" that allows it to attempt to complete tasks even when faced with obstacles, such as pop-up windows [20]. Group 2: User Interaction and Experience - Users can summon the Doubao assistant using a dedicated physical button, which overlays the current app without interrupting ongoing activities [23][28]. - The assistant can read chat contexts and generate responses, allowing for seamless communication without manual typing [25][33]. - It can also perform scheduled tasks, such as checking trending topics on social media and saving information for later use [18][32]. Group 3: Market Position and Collaboration - ByteDance is collaborating with multiple smartphone manufacturers to integrate the Doubao assistant into their devices, indicating a shift towards partnerships with external AI model providers [4][36]. - The assistant's development reflects a broader industry trend where smartphone companies are seeking to enhance their AI capabilities through collaborations rather than solely relying on in-house development [36][37]. - The assistant's current performance shows room for improvement, particularly in executing tasks more efficiently compared to manual operations [36][37].

Core Viewpoint - Nvidia's third-quarter earnings report exceeded expectations, with revenue of $57.01 billion and net profit of $31.91 billion, reflecting year-on-year growth of 62% and 65% respectively, which may alleviate concerns about AI industry valuations in the stock market [2] Group 1: Financial Performance - Nvidia's Q3 revenue was $57.01 billion, surpassing market expectations of $54.92 billion, and showing a year-on-year increase of 62% [2] - The net profit for the same period was $31.91 billion, marking a significant year-on-year increase of 65% [2] Group 2: Market Dynamics - The current AI boom in the U.S. is largely driven by supply-side investments from major tech companies like Microsoft, Google, and Meta, which are heavily investing in Nvidia's GPUs to build computing power centers [2] - There are concerns that the capital expenditures for AI infrastructure are exceeding current actual demand, drawing parallels to the internet bubble of 2000 [3] Group 3: Technological Evolution - Historical tech revolutions often experience bubbles as a necessary phase, with capital flowing in before technology matures, which can lead to resource misallocation but also provides funding for technological advancements [3] - The accumulation of computing power globally may be a necessary step towards achieving Artificial General Intelligence (AGI) [3] Group 4: Future Challenges - The tech giants are entering a challenging phase where the expectations for technology commercialization must catch up with rising anticipations [4] - Investors are increasingly demanding tangible revenue and profit margins rather than just optimistic future projections, indicating a shift in focus from merely accumulating computing power to demonstrating real profitability [4] Group 5: Valuation Concerns - A potential resolution to the current valuation debate could involve a "time for space" process, where gradual technology application leads to more reasonable valuations, requiring patience from market investors [5]