OpenClaw 是一场「无限进化」： 边界持续扩展，参与者不断被重新定义，胜负不在当下，而在持续迭代与适配世界的能力。 统"。 在这样的转变中，理解一线实践，比抽象判断更为重要。 本次， 锦秋基金联合 AgentGuard、周周黑客松、火山引擎 V-Start、飞书 ，在深圳组织了一场「龙虾义诊大会」，邀请 AI Builder 们带着自己的进入"诊室"，并在真实问题中完成一 轮进化的同步。 问题不再是单点能力的提升，而是谁能构建一个可执行、可扩展、能够连接真实世界的"AI操作系 01 投资人养虾的Aha moment 郑晓超｜锦秋基金合伙人 1. Make Something Agent Want | 过去大家讲说要 make something people want | ，今年以来非常大的转变就是要开始 | make something agent want，and a | | --- | --- | --- | | ll people want is agent. | | | 这就带来传统 GUI 软件正在快速API/ CLi 化，而衡量agent产品的 北极星指标 也在 从 DAU、时长等转变成" ...

机器之心编辑部 近日，AI 领域知名专家 Andrej Karpathy 做客一档播客节目，在长达一个小时的对谈中，他不仅回顾了自己近一年的工作状态，也系统阐述了一个正在迅速成形的 新范式：以 Agent 为核心的软件生产方式重构。 Andrej Karpathy 直言自己「病了」，患上了严重的「AI 精神病」。 从去年 12 月开始，他再也没有手写过一行代码。每天 16 小时，他都在和 Agent 对话，同时并行驱动十多个任务；甚至当 token 没被「用满」时，他会感到不安。 这种对 AI 的高度依赖，他称之为「AI 精神病」（AI psychosis）。 Karpathy 判断，App 终将消失。设备只需开放 API，Agent 会成为新的「操作系统」，把音响、灯光、空调、窗帘、安防全部串联起来，甚至只需三段提示词，就 能在一个 WhatsApp 对话里完成统一控制。 未来的用户将不再是人，而是代表人行动的 Agent。整个软件与商业体系，都必须围绕 Agent 进行重构。 甚至连组织本身，也在被重新定义：一个研究机构，本质上就是一组 markdown 文件 —— 角色、流程、协作方式，全都是「代码」； ...

Core Insights - The article highlights the ongoing transition from "technical experimentation" to "engineering paradigm shift" as large models and AI agents become deeply integrated into production environments [2][3] - It emphasizes the need for a comprehensive understanding of this transformation among developers, industry experts, and business leaders, as well as the importance of establishing engineering standards and safety systems to match the rapid advancements in AI technology [2][3] Group 1: Conference Overview - The "2026 Singularity Intelligent Technology Conference" aims to address how to systematically understand the ongoing transformation and find pathways for adaptation [3] - The conference will explore 12 cutting-edge topics, including multimodal models, AI-native development, and agent systems, to create a forward-looking and practical cognitive map for navigating this "tenfold speed transformation" [5] Group 2: Key Topics and Speakers - The "Evolution of Large Language Model Technology" session will feature top scholars and experts who will construct a new coordinate system for the evolution of large model technology [7] - The "Agent Design Patterns and Deep Water Landing" session will focus on building reliable agents, moving away from "blind box" development [13] - The "OpenClaw Industry Practice" session will provide a complete guide for IT leaders and tech enthusiasts on introducing digital employees and adapting to the OpenClaw framework [17] Group 3: AI Infrastructure and Operations - The "AI Infra Infrastructure and Operations" session will present practical guides for transforming operational systems using agent-based approaches, aimed at infrastructure engineers and system architects [21][24] - The session will include insights on automating operations for multi-GPU clusters and enhancing infrastructure with self-awareness and repair capabilities [24] Group 4: AI Application and Industry Practices - The "AI Native Application Innovation and Development Practice" session will showcase successful AI applications that have achieved significant user engagement and valuation, focusing on engineering practices that led to their success [25] - The "AI + Industry Landing Practices" session will provide methodologies for converting large models into tangible business ROI across various sectors, including e-commerce and finance [29] Group 5: Multimodal and Embodied Intelligence - The "Multimodal and World Models" session will cover the underlying technical secrets from video generation to multimodal document understanding, providing a comprehensive engineering path for deployment [39][41] - The "Embodied Intelligence and Intelligent Hardware" session will offer methodologies for achieving large-scale practical applications in high-risk environments, focusing on visual perception and control [47][51] Group 6: Future of AI - The conference serves as a platform for deep communication in the tech field and aims to promote AI ecosystem integration and industry collaborative innovation [53] - It invites global AI industry participants to capture cutting-edge trends and explore paths for industrial upgrades, contributing to the broader application of AI [53]

Core Viewpoint - The distinction between personal and enterprise use of AI agents is critical, with the former being exploratory and the latter carrying significant responsibility. Errors in personal scenarios can be rectified easily, while mistakes in enterprise contexts can lead to severe consequences such as data loss or breaches [1]. Group 1 - The CEO of Feishu, Xie Xin, emphasizes that running an agent on personal computers differs fundamentally from using agents in a corporate environment [1]. - The potential of agent capabilities is exciting, but the safety standards are crucial for their actual implementation in work scenarios [1]. - The more powerful the agent becomes, the greater the associated risks, highlighting the need for stringent safety measures [1].

Core Insights - The 2026 Singularity Intelligent Technology Conference will take place in Shanghai on April 17-18, organized by CSDN and Singularity Intelligence Research Institute [1] - The conference aims to provide attendees with practical survival guides to thrive in a rapidly evolving technological landscape, focusing on the entire lifecycle of AI technology [2][3] Group 1: Key Topics and Pain Points - The conference will cover various layers of AI technology, including perception, control, decision-making, application, infrastructure, research, and architecture [2] - A significant pain point addressed is the limitations of embodied intelligence in low-light or obstructed environments, which can hinder performance in high-risk industrial scenarios [6] - Solutions presented include multi-modal super perception and data-driven regulatory control loops, with insights from experts on overcoming visual blind spots and enhancing operational efficiency in unmanned machinery [7] Group 2: Business AI Evolution - Traditional business AI often stops at sales predictions, while companies require counterfactual reasoning to understand the impact of pricing changes on competitors [8] - The concept of Agentic Commerce will be explored, focusing on causal modeling practices to create business world models that reflect decision-environment-outcome relationships [8][9] - Attendees will learn about the paradigm shift from prediction-driven AI to decision-driven AI, utilizing game theory and simulation to optimize strategies in multi-agent markets [9] Group 3: AI in Software Development - The conference will address the challenges of coding agents and the need for a shift from single-point assistance to collaborative standards in large development teams [18] - A six-dimensional cognitive architecture for agent design will be introduced, emphasizing the importance of memory, reasoning, and collaboration in building reliable agents [20][21] - The event will feature discussions on how AI can reshape software development practices, with insights from leaders in major tech companies [23] Group 4: Future of AI Infrastructure - The conference will delve into the cost and performance challenges of deploying large models, exploring solutions like inference-free techniques and reconfigurable computing [16][17] - Experts will share practical experiences in building AI infrastructures that can dynamically adapt to evolving AI demands, including the development of a 4K super node solution [17] - The focus will be on achieving a balance between effectiveness, speed, and cost in AI applications [16] Group 5: Collaboration and Networking - The conference will feature over 50 leading technology experts discussing topics such as large language models, multi-modal world models, and AI-native applications [22] - Opportunities for collaboration and knowledge sharing will be emphasized, aiming to create verifiable and reusable engineering experiences in the AI era [27]

Investment Rating - The industry investment rating is "Buy" [2] Core Insights - The commercialization of AI agents is accelerating, with significant potential for market penetration. The demand for agents is driving a structural increase in token consumption, indicating a shift towards more efficient and profitable AI applications [5][13][14] - Kimi's revenue is rapidly growing due to the demand for agents and the international expansion of models. The Kimi K2.5 model has surpassed its entire revenue from 2025 within just twenty days of its release, driven by performance improvements and cost advantages [5][17][18] - The report highlights the competitive advantages of domestic models over international counterparts, particularly in terms of cost-effectiveness and performance in agent applications [5][19][20] Summary by Sections AI Weekly Special: Agent - Agents like OpenClaw are becoming key drivers of token growth, with significant increases in token consumption compared to traditional AI applications. OpenClaw's token consumption reached 6.5 trillion, leading the market [13][14] - The commercial model for agents is transitioning from validation to scaling, with high-value users expected to increase, enhancing the profitability of model companies [14][19] Domestic and International Application Stock Price and Valuation Review - Recent trends show a stabilization in US software indices, with B2B SaaS companies performing well. The report notes a 14.2% increase in Hubspot's stock relative to the Nasdaq index [24][25] - The report also indicates a decline in stock prices for AI model companies in Hong Kong, reflecting a market correction despite previous high valuations [30][28] Investment Recommendations - The report suggests focusing on companies that integrate self-developed models, cloud services, and ecosystems. Short-term attention is recommended for Google, while long-term focus should be on Microsoft, Alibaba, and Tencent [5][19]

Core Viewpoint - The article discusses the transition into the "third era" of AI programming, characterized by agents that can independently complete larger tasks with minimal human intervention [1][3]. Summary by Sections Transition from Tab to Agent - The initial phase of coding involved manual key presses, which was transformed by Tab auto-completion, marking the first era of AI-assisted programming. The introduction of agents allowed developers to interact through a prompt-response cycle, leading to the second era. The current third era features agents capable of working over longer time spans and completing larger tasks independently [3][5][6]. Growth of Agent Usage - As of March 2025, the number of Tab users was approximately 2.5 times that of Agent users. However, this ratio has reversed, with Agent users now being twice that of Tab users, and Agent usage has increased rapidly [8][11]. Cloud Agents and Artifacts - Cloud agents operate independently in virtual machines, allowing developers to delegate tasks and focus on other activities. These agents autonomously iterate and test, returning comprehensive outputs that include logs and previews, thus enabling the management of multiple agents simultaneously [13][14]. Internal Changes at Cursor - Within Cursor, 35% of merged code submissions are generated by cloud-based agents. Developers adopting this new workflow typically focus on problem definition and output review rather than step-by-step guidance [15][17]. Future Considerations - There is a recognition that significant work remains to standardize this new development paradigm. Ensuring agents operate efficiently and have access to necessary tools and context is crucial for broader adoption [16]. The recent updates to Cursor have enhanced agent capabilities, allowing for seamless modifications and improved user experience [16]. Community Perspectives - Some community members suggest that the evolution from Tab to synchronous agents and then to cloud agents is an optimization within the same paradigm, emphasizing that the next leap should involve removing the concept of "source code" entirely [18]. Others highlight the need for robust validation mechanisms as autonomous systems scale up code submissions [18].

Core Insights - The AI industry is shifting focus from "can it be done" to "under what conditions, at what cost, and for whom does it create value" as of early 2026 [2][4][6] - Reports from Stanford HAI and other institutions indicate that 2026 will mark a transition from evangelism to evaluation in AI [2][7] Group 1: Investment and ROI - Many companies have completed their first round of generative AI deployment and are beginning to assess their investments and returns [4][5] - A report by Google Cloud titled "The ROI of AI 2025" surveyed 3,466 executives from companies with revenues over $10 million, revealing that sustainable returns come from a system-level implementation of "Agent + Process + Organization" rather than isolated generative AI capabilities [6][29] - Approximately 88% of early adopters of Agentic AI have seen positive returns in at least one generative AI scenario, with the success linked to clear C-level strategies and organizational alignment [30][31] Group 2: Evaluation Standards and AGI - The traditional Scaling Law, which posits that larger models and more data lead to better performance, is becoming inadequate as AI enters high-risk fields like law and medicine [9][10] - There is a growing consensus that the evaluation of AI must evolve to account for the complexity of real-world applications, moving beyond mere capability assessments [10][21] - Wu Enda's proposal for a new Turing-AGI test aims to redefine the standards for evaluating AI, focusing on its ability to perform tasks in unpredictable environments rather than just solving predefined problems [14][19] Group 3: Agentic AI and System Integration - The current focus in AI has shifted from merely enhancing model strength to effectively integrating these models into operational systems [31][32] - Google Cloud's report emphasizes that successful AI implementations are characterized by clear processes and the deployment of Agents in production environments, with over 52% of companies using Agents [33][34] - The report categorizes Agents into three levels, with Level 2 Agents being capable of understanding goals and completing tasks within a defined process, while Level 3 involves collaborative workflows among multiple Agents [37][40] Group 4: Future Directions and Challenges - The future of AI will not be about simply increasing the number of Agents but rather about managing them effectively to ensure stable collaboration and clear accountability [40][41] - The concept of "Skill" is emerging as a critical component in AI, where each task is broken down into manageable, verifiable units that can be monitored and reused [43][44] - The industry is warned about the potential bubble in AI investments, with calls for more empirical research to clarify what AI can and cannot do [27][28]

Core Insights - The article emphasizes that companies are at a crossroads in the AI era, requiring a fundamental shift in their understanding and approach to technology and decision-making [1][5][17] Group 1: Understanding Software - In the digital age, software was viewed merely as a tool for efficiency, but in the AI era, it is evolving into a coded expression of the laws governing operations [2][6] - The path to understanding the world is being rewritten from recognizing patterns to coding them into software, which continuously learns and evolves [2][3] Group 2: Transition to AI-Native Systems - Companies must recognize that digital systems are not equivalent to AI systems, necessitating a comprehensive upgrade to AI-native architectures rather than simply adding AI modules to existing systems [6][12] - The six major transformations in system capabilities highlight the need for a complete migration to AI-native logic [6] Group 3: Misconceptions about AI - A critical misconception is that AI is merely an extension of the internet era, which poses significant risks for companies that fail to adapt their understanding [7][14] - The essence of AI is to complete complex, high-value tasks for specific users, indicating that AI products should be closely tied to business outcomes rather than just traffic generation [9][14] Group 4: Redefining Human-Software Interaction - AI is changing the interaction paradigm, with natural language becoming the new "source code," leading to a restructured relationship between humans and software [10][11] - The emergence of intent-understanding operating systems signifies a shift in how software interprets user needs, moving beyond traditional command execution [10] Group 5: Distinction Between Digitalization and Intelligence - Many companies may appear data-driven but lack true intelligence, as the absence of models in their systems reduces them to mere information processes [12][16] - The fundamental difference lies in the ability to model complex patterns, which is essential for competitive advantage in the AI era [12][18] Group 6: AI's Role in Business Value - The core logic of business is shifting from traffic monetization to value being directly linked to results, with a focus on outcome-based payment models [14][15] - Companies need to prepare for a future where software is atomized into agents that collaborate dynamically to achieve tasks, emphasizing the importance of decision-making over mere task execution [15][16]

Core Insights - The article discusses the emerging trend of AI Agents, which are expected to surpass ChatBots as the primary application form in various fields due to their ability to enhance productivity significantly. Group 1: AI Agents vs. ChatBots - AI Agents can complete entire workflows and deliver results directly, unlike ChatBots, which assist with specific tasks within a workflow [1] - Agents can work in parallel, allowing experienced professionals to collaborate with multiple Agents simultaneously, greatly increasing efficiency [1] - The infrastructure for Agents is still in its infancy, lacking the necessary technology paradigm to support their operational needs [1] Group 2: Daytona's Innovations - Daytona has developed a new type of "composable computer" or "AI sandbox" that allows Agents to run code and manage computer operations with full control over the underlying environment [2] - Daytona recently secured $24 million in Series A funding, led by FirstMark, with participation from several other investors [2] - The founding team of Daytona has a history of creating developer tools and has pivoted from serving human developers to focusing on AI Agents [6][4] Group 3: Technical Specifications - Daytona's infrastructure is designed for speed and concurrency, achieving cold starts in under 60 milliseconds [8] - The system is built entirely in-house, tailored specifically for AI Agents, and does not rely on existing orchestration systems like Kubernetes [9] - Daytona's technology includes strict security boundaries, resource management, and observability, essential for the effective operation of AI Agents [9] Group 4: Market Potential and Future Outlook - The trend of Agentic AI is becoming increasingly prominent, with predictions that Agents will become a significant part of the workforce [17] - The market for Agent-based computing is expected to surpass human-centered computing markets due to the ability of one person to manage multiple Agents [18] - There is a substantial opportunity for entrepreneurs in this space, as the market potential is vast and competition is relatively low [19][20]