如今， 快手 KwaiKAT 团队与南京大学刘佳恒老师 NJU-LINK 实验室 、张煜群教授实验室 ARiSE 合作重磅推出 HiPO（Hybrid Policy Optimization）框架，为 LLM 装上了智能的「思考开关」。 该框架通过创新的混合数据冷启动与混合强化学习奖励系统，使模型能够自主、动态地决策何时该启动详细推理（Think- on），何时该直接给出答案（Think-off）。 当用户向大语言模型提出一个简单问题，比如「单词 HiPPO 里有几个字母 P？」，它却正襟危坐，开始生成一段冗长的推理链：「首先，让我们分析 HiPPO 这个 词，河马的英文单词为 hippo，河马是一种半水生哺乳动物，这里用户用了大写字母，可能有特殊的含义，对于单词 HiPPO，我们可以将其拆分为 H-i-P-P-O，字 母 P 出现在第 3 与第 4 个位置，因此有 2 个字母 P... 让我们简化问题，HiPO 可以拆分为...」 面对这样的「严谨」，用户难免哭笑不得，既浪费了计算资源，也增加了等待时间，甚至更坏的情况是模型被自己冗长的推理链「绕晕了过去」，最终给出了错 误的答案，用户只得捶胸顿足地大喊：「 ...

责任编辑：宋雅芳 新浪科技讯 11月3日上午消息，今日，美团开源全模态模型LongCat-Flash-Omni，官方App同步上线公 测，可体验模型的联网搜索、语音通话等功能。 据悉，新模型是业界首个实现"全模态覆盖、端到端架构、大参数量高效推理"于一体的开源大语言模 型，在开源范畴内实现了对标闭源模型的全模态能力，并凭借创新的架构设计与工程优化，让大参数模 型在多模态任务中实现毫秒级响应。 ...

Core Insights - The article discusses the valuation of the global developer community at $3 trillion, equating it to the GDP of France, highlighting the potential of AI programming to disrupt traditional production relationships and unlock significant value [1][6][5] - It raises concerns about the oversimplification of human creativity into monetary value and the implications of such a perspective on the future of developers [2][3] - The emergence of AI programming as a large-scale application market is emphasized, with significant investments flowing into this sector [6][18] Group 1: AI Programming and Economic Impact - The global developer community, estimated at 30 million, could generate approximately $3 trillion in value, assuming each developer creates $100,000 in value [1][6] - This valuation is comparable to the GDP of France, indicating the substantial economic impact of AI programming [1][6] - The article suggests that AI programming is the first true large-scale application of artificial intelligence, with the potential to create immense value [6][18] Group 2: Disruption of Traditional Software Development - The article posits that traditional computer science education may become obsolete as AI tools evolve, changing the landscape of software development [1][8] - AI tools are increasingly integrated into development processes, leading to unprecedented revenue growth in the IT startup sector [8][12] - The role of developers is expected to shift significantly, with AI taking over many coding tasks, thus altering the traditional software development lifecycle [8][10] Group 3: Future of Development Processes - The development cycle is anticipated to change, with AI agents taking on more responsibilities, potentially reducing the need for human oversight in certain tasks [10][11] - The article discusses the evolving nature of code review, suggesting that AI could handle many aspects of this process, allowing developers to focus on higher-level planning and design [10][14] - The emergence of multi-agent systems in coding could lead to new efficiencies and capabilities in software development [16][20] Group 4: Investment Opportunities and Startup Ecosystem - The article highlights the current environment as an ideal time for launching developer-focused startups, given the significant disruptions in the industry [24][25] - It emphasizes that innovative ideas often come from entrepreneurs rather than investors, suggesting a fertile ground for new ventures in AI programming [24][25] - The potential for creating products specifically for AI agents is identified as a promising area for future startups [25][24]

一家早期机构的转型升级。 这是极具风向标的一幕。 作者 I 吴琼 报道 I 投资界PEdaily 投资界获悉，英诺天使基金进行一次关键升级——内部将正式分为英诺天使基金、英诺科创基金两个品牌，各自组成10到15人的团 队，覆盖不同项目需求，专注早期科技投资。 这是英诺内部一次主动进化。早在2 01 9年，英诺就设立了科创基金，聚焦科技。经过六年探索，未来英诺科创基金将作为独立品牌运 作。一个鲜明特点是，升级后英诺科创基金将在投资金额上出手更大。 "投早投小投科技"已经成为共识，英诺此时动作，恰是中国早期创投生态变迁的一缕缩影。 英诺之变 两个团队，独立运作 李竹向投资界透露，这场转型酝酿六年之久。 在他的记忆里，2 018年底英诺内部曾进行过一次复盘。彼时团队发现，梳理前五年投过的项目，当中科技项目带来的回报普遍更高。 加上团队成员多是清华等高校的理工科背景出身，因此内部将重点转移到科技投资上。 2019年，英诺成立英诺科创基金一期，规模3.6亿元。在此之前，英诺作为一家天使投资机构几乎什么都投，投资版图遍布科技、互联 网、文娱、消费……而英诺科创基金的目标明确——只投科技。 亲历这几年市场动荡，英诺决心以更 ...

henry 发自 凹非寺 量子位 | 公众号 QbitAI AI Coding火了大半年，AI Debugging也来了！ 刚刚，OpenAI发布由GPT-5驱动的"白帽"Agent—— Aardvark（土豚） 。 这只"AI安全研究员"能帮助开发者和安全团队， 在大规模代码库中自动发现并修复安全漏洞 。 据OpenAI报告，Aardvark已识别出了 92% 的已知与人工注入漏洞，而且能定位仅在复杂条件下出现的问题。 OpenAI副总裁 Matt Knight 表示： 我们的开发者发现，土豚在清晰地解释问题并引导他们找到修复方案方面确实非常有价值。这个信号告诉我们，我们正走在一条有意义 的道路上。 而且，不仅OpenAI。 整个10月 Anthropic 、 谷歌 、 微软 基本上是前脚跟后脚发布了类似的白帽Agent。 Agentic AI +自动修补漏洞 OpenAI对这款白帽Aardvark的官方描述是—— 代理型安全研究员 （agentic security researcher） Aardvark的核心任务是持续分析源代码仓库，以识别安全漏洞、评估可利用性、确定风险等级，并提出有针对性的修复方案 ...

这些被称为"工业智能体"的系统，不仅需要具备自主推理、规划与工具使用能力，更要适配复杂业务逻辑、严苛安全标准与领域知识壁垒——如何 将通用智能体的技术潜力，转化为驱动产业变革的实际生产力，成为当前AI落地的核心挑战。 点击下方 卡片 ，关注" 大模型之心Tech "公众号 戳我-> 领取大模型巨卷干货 本文只做学术分享，如有侵权，联系删文 随着大语言模型（LLMs）能力的爆发式增长，AI智能体 (Agent) 已从通用场景探索，逐步深入到金融、医疗、制造等知识密集、高风险的工业领 域。 近期，来自哈工深与华为的研究团队系统梳理了LLM驱动工业智能体的技术演进、应用实践与评测体系，发布了 一篇覆盖300+篇研究的综述 《Empowering Real-World: A Survey on the Technology, Practice, and Evaluation of LLM-driven Industry Agents》。本文创新性提出"能力成熟度框架"， 从"流程执行系统"到"自适应社会系统"，清晰勾勒出工业智能体的进化路径，为科研与产业落地提供了完整参考蓝图。 论文地址 ：https://arxiv ...

这一成果迅速在产业界引发热烈讨论。 日 前 ， DeepSeek 在 人 工 智 能 开 源 社 区 Hugging Face 上 发 布 了 一 个 全 新 的 多 模 态 模 型 DeepSeek-OCR 。 在华为旗下的学术平台"黄大年茶思屋"上，有技术专家甚至指出，该模型的核心构件视觉 encoder的高效解码，为光计算和量子计算在LLM（注：大语言模型）领域的引入提供了明确 的技术路径。 10月29日，图灵量子相关负责人在接受21世纪经济报道记者采访时表示， DeepSeek-OCR技 术能更有效地将光计算高并行性和低功耗优势发挥出来，相信很快便会有光计算芯片结合大 模型的应用出现 。 光学压缩破局 一直以来，上下文的长度是困扰大模型性能的重要瓶颈 。比如，上下文窗口过小，会导致模 型无法一次性阅读用户之前的输入信息（比如文章），影响推理的准确性。 针 对 这 个 痛 点 ， 业 内 提 出 了 稀 疏 注 意 力 、 检 索 增 强 生 成 等 多 种 技 术 来 应 对 。 这 一 次 ， DeepSeek首次提出"上下文光学压缩"（Contexts Optical Compression）技术 ...

Core Viewpoint - Nvidia's market capitalization has officially surpassed $5 trillion, making it the first company in the world to reach this milestone, showcasing unprecedented growth speed and market influence [1][2]. Market Performance - Nvidia's stock price increased by approximately 3% to $207.16, resulting in a market cap of $5.03 trillion [2]. - Over the past six months, Nvidia's stock price has surged by about 90%, exceeding the combined market capitalization of major indices in Germany, France, and Italy [5]. - Nvidia's market cap surpasses the total market capitalization of competitors such as AMD, Arm, ASML, Broadcom, Intel, Lam Research, Qualcomm, and TSMC, as well as entire sectors like utilities, industrials, and consumer staples within the S&P 500 [4]. Growth Trajectory - Nvidia's market value was around $400 billion three years ago, prior to the launch of generative AI tools like ChatGPT. Following ChatGPT's release, Nvidia's market cap quickly exceeded $1 trillion [9]. - The growth trajectory of Nvidia has outpaced that of tech giants Apple and Microsoft, which recently reached a market cap of $4 trillion [11]. Demand and Orders - Nvidia's GPUs are considered the driving force behind the entire AI industry, with strong demand reflected in order data. The company has shipped 6 million units of its Blackwell chip and has 14 million units on order [12][13]. - Nvidia's CEO, Jensen Huang, has made optimistic sales forecasts, predicting chip sales to exceed $300 billion in 2026, significantly higher than Wall Street's average expectation of $258 billion [14]. Industry Investment - The substantial demand for Nvidia's products primarily comes from large tech companies investing heavily in data center infrastructure necessary for running AI models [15]. Valuation Concerns - Despite the impressive stock performance, there are concerns about a potential bubble, with some analysts comparing the current AI stock surge to the internet bubble of the early 2000s. Companies are incurring significant debt while generating relatively low revenue [16]. - Nvidia's valuation is under scrutiny, with its stock trading at approximately 33 times its expected earnings for the next year, compared to an average P/E ratio of 24 for the S&P 500 [16].

Core Viewpoint - The article discusses the evolving landscape of AI investment in China, highlighting the shift from merely "catching up" to establishing a unique innovation path driven by domestic capabilities and market conditions [6][11]. Group 1: Investment Trends - BlueRun Ventures has been actively investing in various AI sectors, including foundational models, embodied intelligence, and AI hardware, creating a systematic investment map [5][14]. - The firm emphasizes the importance of open-source models and their cost-effectiveness, which fosters rapid iteration and application development [9][10]. - The investment strategy is centered around five key trends, including the rise of open-source large language models, reinforcement learning, and the development of autonomous systems [9][10]. Group 2: Market Dynamics - China's economic structure is undergoing a transformation, with technology-driven growth becoming the new mainline, supported by increasing domestic demand and consumption [7][8]. - The competition between Chinese AI entrepreneurs and their U.S. counterparts is characterized by a dual-track approach, leveraging open-source ecosystems and diverse application scenarios [7][8]. - The emergence of successful Chinese AI products, such as DeepSeek, signifies a shift towards independent innovation and global competitiveness [8][11]. Group 3: Talent and Ecosystem - The density of talent, particularly in AI and related fields, is crucial for the success of new ventures, with a notable influx of young, highly educated entrepreneurs returning to China [13][16]. - BlueRun Ventures has established a supportive ecosystem for entrepreneurs, including initiatives like Boomi ng Camp and Boomi ng Hub, to foster collaboration and innovation [18][19]. - The firm believes that the future of AI investment lies in early-stage opportunities, emphasizing the importance of independent thinking amidst market noise [19][20].

Core Insights - DeepSeek has launched a new multimodal model, DeepSeek-OCR, which has sparked significant discussion in the industry regarding its potential applications in AI and quantum computing [1] - The model's visual encoder is noted for its efficient decoding capabilities, providing a clear technical pathway for integrating optical and quantum computing into large language models (LLMs) [1][2] Group 1: Technological Innovations - DeepSeek-OCR introduces "Contexts Optical Compression," allowing text to be processed as images, theoretically enabling infinite context and achieving a token compression of 7-20 times [2][3] - The model maintains 97% decoding accuracy at 10x compression and 60% accuracy at 20x compression, which is crucial for implementing memory and forgetting mechanisms in LLMs [2][3] Group 2: Implications for Optical Computing - The technology reduces the number of data segmentation and assembly operations, thereby lowering overall computational load and pressure on backend hardware [3][4] - DeepSeek-OCR's approach may facilitate the integration of optical computing chips with large models, leveraging the high parallelism and low power consumption of optical technologies [3][4] Group 3: Industry Challenges and Developments - Current challenges for optical computing include the need for advanced photonic-electronic integration and a mature software ecosystem to support large-scale development [5] - Key players in the optical computing space include domestic companies like Turing Quantum and international firms such as Lightmatter and Cerebras Systems, with Turing Quantum making strides in thin-film lithium niobate technology [5]