{
  "title": "2026-06-04 AI Daily | Codex's All-Out Offensive, AI Programming Enters the New 'Delivery as Finished Product' Stage",
  "url": "https://miaok.ong/en/ai-daily/ai-daily-2026-06-04/",
  "date": "2026-06-04T07:00:00+08:00",
  "lastmod": "2026-06-04T07:00:00+08:00",
  "type": "ai-daily",
  "kind": "page",
  "language": "en",
  "description": "OpenAI Codex, with its \u0026lsquo;Sites\u0026rsquo; feature, enables single-sentence website building and remote cross-machine collaboration, surpassing 5 million weekly active users. Meanwhile, Anthropic actively promotes Claude Code with free courses, and the community is shifting from tool comparison to workflow integration. On-device models have achieved production usability on consumer-grade hardware, and Agents are progressing from code generation to autonomous delivery, with product path divergence and engineering implementation simultaneously accelerating.",
  "keywords": null,
  "tags": [],
  "categories": [],
  "author": "Mark (Miao) Kong",
  "image": "https://miaok.ong/images/avatar.jpg",
  "content": "\u003ch1 id=\"2026-06-04-ai-daily--codexs-full-scale-push-as-ai-programming-enters-a-new-ready-to-deploy-stage\"\u003e\n  2026-06-04 AI Daily | Codex\u0026rsquo;s Full-Scale Push, as AI Programming Enters a New \u0026ldquo;Ready-to-Deploy\u0026rdquo; Stage\n  \u003ca class=\"heading-link\" href=\"#2026-06-04-ai-daily--codexs-full-scale-push-as-ai-programming-enters-a-new-ready-to-deploy-stage\"\u003e\n    \u003ci class=\"fa-solid fa-link\" aria-hidden=\"true\" title=\"Link to heading\"\u003e\u003c/i\u003e\n    \u003cspan class=\"sr-only\"\u003eLink to heading\u003c/span\u003e\n  \u003c/a\u003e\n\u003c/h1\u003e\n\u003cblockquote\u003e\n\u003cp\u003eWith its \u0026ldquo;Sites\u0026rdquo; feature, OpenAI Codex enables one-sentence website creation and remote cross-machine collaboration, surpassing 5 million weekly active users. Simultaneously, Anthropic is heavily promoting Claude Code with free courses, shifting the community\u0026rsquo;s focus from tool comparison to workflow integration. On-device models have achieved production-level usability on consumer hardware, and agents are evolving from code generation to autonomous delivery, accelerating both the divergence of product paths and their engineering implementation.\u003c/p\u003e\n\u003c/blockquote\u003e\n\u003ch2 id=\"-in-depth-guide-to-this-issues-watch-list\"\u003e\n  📖 In-depth Guide to This Issue\u0026rsquo;s Watch List\n  \u003ca class=\"heading-link\" href=\"#-in-depth-guide-to-this-issues-watch-list\"\u003e\n    \u003ci class=\"fa-solid fa-link\" aria-hidden=\"true\" title=\"Link to heading\"\u003e\u003c/i\u003e\n    \u003cspan class=\"sr-only\"\u003eLink to heading\u003c/span\u003e\n  \u003c/a\u003e\n\u003c/h2\u003e\n\u003cp\u003eToday, OpenAI released two major policy documents—from a public policy agenda to a federal blueprint for advanced AI governance in the US—clearly outlining its core principles on security, workforce transition, and global standards. This is highly recommended reading for all teams focused on compliance and engineering implementation. In the life sciences, the latest capability upgrades for GPT-Rosalind and the debut of the \u0026ldquo;AI Co-scientist\u0026rdquo; are pushing complex analyses like medicinal chemistry and genomics into a new phase and are well worth watching. The engineering side is equally exciting: a YC partner\u0026rsquo;s in-depth conversation with the founder of OpenClaw breaks down how to build a truly AI-native service company, while the Wasmer team leveraged Codex to cut their Node.js edge runtime development from one year to just two weeks, making it a textbook case for AI\u0026rsquo;s role in rebuilding the developer toolchain.\u003c/p\u003e\n\u003ch2 id=\"-trending-ai-news-on-x\"\u003e\n  🌐 Trending AI News on X\n  \u003ca class=\"heading-link\" href=\"#-trending-ai-news-on-x\"\u003e\n    \u003ci class=\"fa-solid fa-link\" aria-hidden=\"true\" title=\"Link to heading\"\u003e\u003c/i\u003e\n    \u003cspan class=\"sr-only\"\u003eLink to heading\u003c/span\u003e\n  \u003c/a\u003e\n\u003c/h2\u003e\n\u003ch3 id=\"topic-1-openclaw-202661-brings-native-windows-support-and-self-improving-agents\"\u003e\n  Topic 1: OpenClaw 2026.6.1 Brings Native Windows Support and Self-Improving Agents\n  \u003ca class=\"heading-link\" href=\"#topic-1-openclaw-202661-brings-native-windows-support-and-self-improving-agents\"\u003e\n    \u003ci class=\"fa-solid fa-link\" aria-hidden=\"true\" title=\"Link to heading\"\u003e\u003c/i\u003e\n    \u003cspan class=\"sr-only\"\u003eLink to heading\u003c/span\u003e\n  \u003c/a\u003e\n\u003c/h3\u003e\n\u003cul\u003e\n\u003cli\u003eCategory: AI · News\u003c/li\u003e\n\u003cli\u003eSummary: Trending since: 2 hours ago, Related posts: 204\u003c/li\u003e\n\u003cli\u003eWhat it is: The open-source project OpenClaw released version 2026.6.1, introducing native Windows support and integrating self-improving AI agents for the first time.\u003c/li\u003e\n\u003cli\u003eWhy it matters: Native Windows support removes a major barrier to cross-platform deployment, while the self-improving agents demonstrate a viable path for AI systems to achieve autonomous iteration within practical tools, potentially driving the developer ecosystem toward a more open and automated evolution.\u003c/li\u003e\n\u003cli\u003eDiscussion summary: Discussions are focused on whether the Windows version can match the performance of its Linux counterpart, the safety boundaries for self-improving agents, and whether this capability will further diminish the role of human developers in tool maintenance.\u003c/li\u003e\n\u003c/ul\u003e\n\u003ch3 id=\"topic-2-anthropic-engineers-share-free-claude-code-mastery-talks\"\u003e\n  Topic 2: Anthropic Engineers Share Free Claude Code Mastery Talks\n  \u003ca class=\"heading-link\" href=\"#topic-2-anthropic-engineers-share-free-claude-code-mastery-talks\"\u003e\n    \u003ci class=\"fa-solid fa-link\" aria-hidden=\"true\" title=\"Link to heading\"\u003e\u003c/i\u003e\n    \u003cspan class=\"sr-only\"\u003eLink to heading\u003c/span\u003e\n  \u003c/a\u003e\n\u003c/h3\u003e\n\u003cul\u003e\n\u003cli\u003eCategory: AI · News\u003c/li\u003e\n\u003cli\u003eSummary: Trending since: 5 hours ago, Related posts: 391\u003c/li\u003e\n\u003cli\u003eWhat it is: Anthropic engineers have released a series of free \u0026ldquo;Claude Code Mastery\u0026rdquo; talks.\u003c/li\u003e\n\u003cli\u003eWhy it matters: This indicates that Anthropic is directly empowering the developer community through official tech talks to accelerate the adoption of its AI programming tool, Claude Code, and challenge existing competitors, potentially impacting the developer tool ecosystem.\u003c/li\u003e\n\u003cli\u003eDiscussion summary: Discussions on X are centered on the content and practical tips from the talks, as well as debates comparing the capabilities, pricing, and developer experience of Claude Code versus competitors like GitHub Copilot and Cursor.\u003c/li\u003e\n\u003c/ul\u003e\n\u003ch3 id=\"topic-3-openais-its-time-to-fly-video-spotlights-codex-coding-power\"\u003e\n  Topic 3: OpenAI\u0026rsquo;s \u0026lsquo;It\u0026rsquo;s Time to Fly\u0026rsquo; Video Spotlights Codex Coding Power\n  \u003ca class=\"heading-link\" href=\"#topic-3-openais-its-time-to-fly-video-spotlights-codex-coding-power\"\u003e\n    \u003ci class=\"fa-solid fa-link\" aria-hidden=\"true\" title=\"Link to heading\"\u003e\u003c/i\u003e\n    \u003cspan class=\"sr-only\"\u003eLink to heading\u003c/span\u003e\n  \u003c/a\u003e\n\u003c/h3\u003e\n\u003cul\u003e\n\u003cli\u003eCategory: AI · News\u003c/li\u003e\n\u003cli\u003eSummary: Trending since: , Related posts: 991\u003c/li\u003e\n\u003cli\u003eWhat it is: OpenAI released an \u0026ldquo;It\u0026rsquo;s Time to Fly\u0026rdquo; video showcasing the Codex model\u0026rsquo;s ability to convert natural language instructions into runnable code in real-time.\u003c/li\u003e\n\u003cli\u003eWhy it matters: The demo visually represents the evolution of code generation from an auxiliary tool to an autonomous development agent, signaling a lower barrier to entry for programming, a new paradigm for human-computer collaboration, and profound implications for the software engineering process.\u003c/li\u003e\n\u003cli\u003eDiscussion summary: The focus is on Codex\u0026rsquo;s impact on developer careers, the reliability and security of the generated code, and whether OpenAI\u0026rsquo;s packaging of the demo as an emotive brand campaign excessively downplays the technology\u0026rsquo;s limitations.\u003c/li\u003e\n\u003c/ul\u003e\n\u003ch3 id=\"topic-4-openai-codex-team-aims-for-ultimate-desktop-coding-app\"\u003e\n  Topic 4: OpenAI Codex Team Aims for Ultimate Desktop Coding App\n  \u003ca class=\"heading-link\" href=\"#topic-4-openai-codex-team-aims-for-ultimate-desktop-coding-app\"\u003e\n    \u003ci class=\"fa-solid fa-link\" aria-hidden=\"true\" title=\"Link to heading\"\u003e\u003c/i\u003e\n    \u003cspan class=\"sr-only\"\u003eLink to heading\u003c/span\u003e\n  \u003c/a\u003e\n\u003c/h3\u003e\n\u003cul\u003e\n\u003cli\u003eCategory: AI · News\u003c/li\u003e\n\u003cli\u003eSummary: Trending since: 18 hours ago, Related posts: 473\u003c/li\u003e\n\u003cli\u003eWhat it is: OpenAI\u0026rsquo;s Codex team is developing what it positions as the \u0026ldquo;ultimate\u0026rdquo; desktop programming application.\u003c/li\u003e\n\u003cli\u003eWhy it matters: This move signifies OpenAI\u0026rsquo;s expansion of AI coding capabilities from cloud-based IDEs to a standalone desktop app, attempting to redefine the localized workflow for developer-AI collaboration. It could trigger an escalation in competition within the coding tool ecosystem.\u003c/li\u003e\n\u003cli\u003eDiscussion summary: Discussions on X focus on the app\u0026rsquo;s differentiating advantages over existing tools like GitHub Copilot and Cursor. There is speculation about its potential for deep code understanding and autonomous programming capabilities, along with divided opinions on its product maturity, business model, and potential impact on developer employment.\u003c/li\u003e\n\u003c/ul\u003e\n\u003ch3 id=\"topic-5-reve-20-launches-as-top-ai-image-generator-with-layout-control\"\u003e\n  Topic 5: Reve 2.0 Launches as Top AI Image Generator with Layout Control\n  \u003ca class=\"heading-link\" href=\"#topic-5-reve-20-launches-as-top-ai-image-generator-with-layout-control\"\u003e\n    \u003ci class=\"fa-solid fa-link\" aria-hidden=\"true\" title=\"Link to heading\"\u003e\u003c/i\u003e\n    \u003cspan class=\"sr-only\"\u003eLink to heading\u003c/span\u003e\n  \u003c/a\u003e\n\u003c/h3\u003e\n\u003cul\u003e\n\u003cli\u003eCategory: AI · News\u003c/li\u003e\n\u003cli\u003eOverview: Trending since: 5 hours ago, Related posts: 1100\u003c/li\u003e\n\u003cli\u003eWhat it is: AI image generation model Reve 2.0 has been officially released, featuring advanced layout control capabilities.\u003c/li\u003e\n\u003cli\u003eWhy it matters: Precise layout control has always been a weak point for diffusion models in image generation. Reve 2.0 elevates this to a new level, promising to shift AI creation from random aesthetics to controllable design, thus lowering the barrier for commercial adoption.\u003c/li\u003e\n\u003cli\u003eDiscussion summary: The discussion focuses on the real-world usability of its layout control, comparisons with competitors like Midjourney, and whether it will have an open API or be open-sourced. General users are more concerned with generation speed and cost.\u003c/li\u003e\n\u003c/ul\u003e\n\u003ch3 id=\"topic-6-openais-codex-reaches-5-million-weekly-users-as-productivity-tool\"\u003e\n  Topic 6: OpenAI\u0026rsquo;s Codex Reaches 5 Million Weekly Users as Productivity Tool\n  \u003ca class=\"heading-link\" href=\"#topic-6-openais-codex-reaches-5-million-weekly-users-as-productivity-tool\"\u003e\n    \u003ci class=\"fa-solid fa-link\" aria-hidden=\"true\" title=\"Link to heading\"\u003e\u003c/i\u003e\n    \u003cspan class=\"sr-only\"\u003eLink to heading\u003c/span\u003e\n  \u003c/a\u003e\n\u003c/h3\u003e\n\u003cul\u003e\n\u003cli\u003eCategory: AI · Sports\u003c/li\u003e\n\u003cli\u003eOverview: Trending since: 1 day ago, Related posts: 34,000\u003c/li\u003e\n\u003cli\u003eWhat it is: OpenAI\u0026rsquo;s programming assistant, Codex, has surpassed 5 million weekly active users, reflecting its rapid adoption as a productivity tool.\u003c/li\u003e\n\u003cli\u003eWhy it matters: This milestone marks the transition of AI-assisted development from an early-adopter phase to large-scale application. It helps improve the efficiency and raise the bar for software engineering as a whole, while also highlighting the potential of code generation models for commercialization and user retention.\u003c/li\u003e\n\u003cli\u003eDiscussion summary: The focus is on the actual extent to which Codex improves development efficiency, its differentiating advantages over GitHub Copilot, and concerns from some developers that over-reliance on AI could weaken fundamental programming skills. There is also discussion about its potential applications in non-traditional programming scenarios, such as sports data analysis.\u003c/li\u003e\n\u003c/ul\u003e\n\u003ch3 id=\"topic-7-nvidias-jensen-huang-shares-beer-on-computex-floor-with-gigabyte-exec\"\u003e\n  Topic 7: NVIDIA\u0026rsquo;s Jensen Huang Shares Beer on Computex Floor with Gigabyte Exec\n  \u003ca class=\"heading-link\" href=\"#topic-7-nvidias-jensen-huang-shares-beer-on-computex-floor-with-gigabyte-exec\"\u003e\n    \u003ci class=\"fa-solid fa-link\" aria-hidden=\"true\" title=\"Link to heading\"\u003e\u003c/i\u003e\n    \u003cspan class=\"sr-only\"\u003eLink to heading\u003c/span\u003e\n  \u003c/a\u003e\n\u003c/h3\u003e\n\u003cul\u003e\n\u003cli\u003eCategory: AI · News\u003c/li\u003e\n\u003cli\u003eOverview: Trending since: , Related posts: 472\u003c/li\u003e\n\u003cli\u003eWhat it is: NVIDIA CEO Jensen Huang shared a beer with a Gigabyte executive on the Computex show floor.\u003c/li\u003e\n\u003cli\u003eWhy it matters: The interaction highlights the close relationship between NVIDIA and its key hardware partners, signaling continued collaboration in GPU supply and system integration within the AI computing ecosystem. This has a direct impact on the large-scale deployment of AI infrastructure and industry confidence.\u003c/li\u003e\n\u003cli\u003eDiscussion summary: On the X platform, the main points of focus are: first, potential details of new products or technical collaborations between the two parties; second, this relaxed atmosphere is seen as a sign of a stable AI computing supply chain. Some opinions also discuss the role of Jensen Huang\u0026rsquo;s personal style in fostering industry relationships and the changing roles of ODM manufacturers like Gigabyte in the AI server market.\u003c/li\u003e\n\u003c/ul\u003e\n\u003ch4 id=\"ai-public-opinion-summary-on-x-today\"\u003e\n  AI Public Opinion Summary on X Today\n  \u003ca class=\"heading-link\" href=\"#ai-public-opinion-summary-on-x-today\"\u003e\n    \u003ci class=\"fa-solid fa-link\" aria-hidden=\"true\" title=\"Link to heading\"\u003e\u003c/i\u003e\n    \u003cspan class=\"sr-only\"\u003eLink to heading\u003c/span\u003e\n  \u003c/a\u003e\n\u003c/h4\u003e\n\u003cp\u003eBuilding on the confidence in computing power collaboration signaled by Jensen Huang and Gigabyte\u0026rsquo;s toast, today\u0026rsquo;s public discourse centers on the collective leap forward of AI programming tools and agents: OpenAI Codex released an emotive demo, hit 5 million weekly active users, and launched its \u0026ldquo;ultimate\u0026rdquo; desktop application; Anthropic is promoting Claude Code with free courses; and the open-source OpenClaw has integrated a self-improving agent on Windows for the first time. Meanwhile, the image generation model Reve 2.0 is pushing creation from random to controllable with its precise layout control. The consensus is that AI has significantly increased development efficiency and lowered the entry barrier, but the community is widely concerned about the reliability of generated code, the safety boundaries of autonomous agents, and the erosion of human skills. Points of disagreement center on the practical capability differences between tools like Codex, GitHub Copilot, Cursor, and Claude Code, the debate between open-source and closed-source approaches, and the real impact of AI on developer employment. Potential risks are becoming clearer: self-improving agents might breach safety constraints, over-reliance is eroding fundamental programming skills, and vendors packaging demos into brand narratives could obscure technical limitations, leading to a brewing crisis of trust.\u003c/p\u003e\n\u003ch2 id=\"-influencer-insights\"\u003e\n  💡 Influencer Insights\n  \u003ca class=\"heading-link\" href=\"#-influencer-insights\"\u003e\n    \u003ci class=\"fa-solid fa-link\" aria-hidden=\"true\" title=\"Link to heading\"\u003e\u003c/i\u003e\n    \u003cspan class=\"sr-only\"\u003eLink to heading\u003c/span\u003e\n  \u003c/a\u003e\n\u003c/h2\u003e\n\u003ch1 id=\"ai-industry-daily-2026-06-0304\"\u003e\n  AI Industry Daily (2026-06-03/04)\n  \u003ca class=\"heading-link\" href=\"#ai-industry-daily-2026-06-0304\"\u003e\n    \u003ci class=\"fa-solid fa-link\" aria-hidden=\"true\" title=\"Link to heading\"\u003e\u003c/i\u003e\n    \u003cspan class=\"sr-only\"\u003eLink to heading\u003c/span\u003e\n  \u003c/a\u003e\n\u003c/h1\u003e\n\u003ch2 id=\"i-widely-followed-tech-trends-and-product-highlights\"\u003e\n  I. Widely Followed Tech Trends and Product Highlights\n  \u003ca class=\"heading-link\" href=\"#i-widely-followed-tech-trends-and-product-highlights\"\u003e\n    \u003ci class=\"fa-solid fa-link\" aria-hidden=\"true\" title=\"Link to heading\"\u003e\u003c/i\u003e\n    \u003cspan class=\"sr-only\"\u003eLink to heading\u003c/span\u003e\n  \u003c/a\u003e\n\u003c/h2\u003e\n\u003ch3 id=\"1-the-codex-ecosystem-continues-to-explode-becoming-the-top-choice-for-agent-development\"\u003e\n  1. \u003cstrong\u003eThe Codex ecosystem continues to explode, becoming the top choice for Agent development\u003c/strong\u003e\n  \u003ca class=\"heading-link\" href=\"#1-the-codex-ecosystem-continues-to-explode-becoming-the-top-choice-for-agent-development\"\u003e\n    \u003ci class=\"fa-solid fa-link\" aria-hidden=\"true\" title=\"Link to heading\"\u003e\u003c/i\u003e\n    \u003cspan class=\"sr-only\"\u003eLink to heading\u003c/span\u003e\n  \u003c/a\u003e\n\u003c/h3\u003e\n\u003cp\u003eIn the last 24 hours, Codex has been the absolute central topic. Multiple bloggers have been intensively discussing its new features and usage tips:\u003c/p\u003e\n\u003cul\u003e\n\u003cli\u003e\u003cstrong\u003eSites Feature Launch\u003c/strong\u003e: @vista8 pointed out that the Codex enterprise edition has added the \u003ccode\u003e@site\u003c/code\u003e command, which can \u0026ldquo;build a website in one sentence\u0026rdquo; and share the URL. He believes, \u0026ldquo;Anthropic only does the Design, OpenAI goes a step further, handling not just design but also website generation.\u0026rdquo;\u003c/li\u003e\n\u003cli\u003e\u003cstrong\u003eIn-depth Use of the Goal Command\u003c/strong\u003e: @vista8 shared a six-element template for the Goal command based on research with GPT 5.5 Pro (Result, Verification, Constraints, Boundaries, Iteration, Blocking Conditions), and recommends first using the \u003ccode\u003e/plan\u003c/code\u003e mode to have the AI ask clarifying questions to narrow down the requirements.\u003c/li\u003e\n\u003cli\u003e\u003cstrong\u003eRemote Collaboration and Automation\u003c/strong\u003e: @vista8 discovered that Codex supports \u0026ldquo;using one computer to remotely command another to write code,\u0026rdquo; achieving cross-machine collaboration through the \u0026ldquo;control other devices\u0026rdquo; feature.\u003c/li\u003e\n\u003cli\u003e\u003cstrong\u003eUsage Refresh Mechanism\u003c/strong\u003e: @dotey retweeted a confirmation that the Codex free and go plans have now switched to a \u003cstrong\u003emonthly usage refresh\u003c/strong\u003e.\u003c/li\u003e\n\u003c/ul\u003e\n\u003ch3 id=\"2-comparison-and-migration-between-claude-code-and-codex\"\u003e\n  2. \u003cstrong\u003eComparison and Migration between Claude Code and Codex\u003c/strong\u003e\n  \u003ca class=\"heading-link\" href=\"#2-comparison-and-migration-between-claude-code-and-codex\"\u003e\n    \u003ci class=\"fa-solid fa-link\" aria-hidden=\"true\" title=\"Link to heading\"\u003e\u003c/i\u003e\n    \u003cspan class=\"sr-only\"\u003eLink to heading\u003c/span\u003e\n  \u003c/a\u003e\n\u003c/h3\u003e\n\u003cul\u003e\n\u003cli\u003e@ruanyf observed that \u0026ldquo;many people have recently jumped ship to Codex, and the reviews are quite good,\u0026rdquo; mentioning that the official Codex plan starts at $20/month.\u003c/li\u003e\n\u003cli\u003e@dotey\u0026rsquo;s personal ranking: \u003cstrong\u003eCodex App \u0026gt; Cursor \u0026gt; Claude Desktop\u003c/strong\u003e\u003c/li\u003e\n\u003cli\u003e@Pluvio9yte shared the simplest solution for integrating DeepSeek with Claude Code, and how to troubleshoot residual configuration issues with cc-switch (\u003ccode\u003e~/.cc-switch/cc-switch.db\u003c/code\u003e)\u003c/li\u003e\n\u003c/ul\u003e\n\u003ch3 id=\"3-leap-in-on-device-model-capabilities-local-deployment-becomes-a-new-trend\"\u003e\n  3. \u003cstrong\u003eLeap in On-Device Model Capabilities, Local Deployment Becomes a New Trend\u003c/strong\u003e\n  \u003ca class=\"heading-link\" href=\"#3-leap-in-on-device-model-capabilities-local-deployment-becomes-a-new-trend\"\u003e\n    \u003ci class=\"fa-solid fa-link\" aria-hidden=\"true\" title=\"Link to heading\"\u003e\u003c/i\u003e\n    \u003cspan class=\"sr-only\"\u003eLink to heading\u003c/span\u003e\n  \u003c/a\u003e\n\u003c/h3\u003e\n\u003cul\u003e\n\u003cli\u003e@zhixianio detailed his \u0026ldquo;ascetic-style\u0026rdquo; experience using local models: Qwen3.6-35B-A3B-oQ6-fp16-mtp paired with oMLX, \u003cstrong\u003ethe response speed is faster than remote LLMs, and the intelligence is on point\u003c/strong\u003e. In PA scenarios, the native multi-modal experience is \u0026ldquo;even better than DSV4 Pro.\u0026rdquo;\u003c/li\u003e\n\u003cli\u003e@zhixianio is also following MiniCPM5-1B (\u0026ldquo;the strongest open-source base model under 2B\u0026rdquo;) and the AMD Ryzen AI Halo on-device platform.\u003c/li\u003e\n\u003c/ul\u003e\n\u003ch3 id=\"4-agent-memory-solutions-and-workflow-optimization\"\u003e\n  4. \u003cstrong\u003eAgent Memory Solutions and Workflow Optimization\u003c/strong\u003e\n  \u003ca class=\"heading-link\" href=\"#4-agent-memory-solutions-and-workflow-optimization\"\u003e\n    \u003ci class=\"fa-solid fa-link\" aria-hidden=\"true\" title=\"Link to heading\"\u003e\u003c/i\u003e\n    \u003cspan class=\"sr-only\"\u003eLink to heading\u003c/span\u003e\n  \u003c/a\u003e\n\u003c/h3\u003e\n\u003cul\u003e\n\u003cli\u003e@Pluvio9yte shared a discussion on \u0026ldquo;the current closest-to-correct solution for Agent memory.\u0026rdquo;\u003c/li\u003e\n\u003cli\u003e@servasyy_ai\u0026rsquo;s breakdown of the Claude Code Workflow was shared by multiple bloggers (@Pluvio9yte commented it was \u0026ldquo;full of substance\u0026rdquo;).\u003c/li\u003e\n\u003cli\u003eThe Chronicle feature mentioned by @jungeAGI (which uses screen context to enhance memory) has garnered attention.\u003c/li\u003e\n\u003c/ul\u003e\n\u003chr\u003e\n\u003ch2 id=\"ii-unique-perspectives-and-industry-foresight\"\u003e\n  II. Unique Perspectives and Industry Foresight\n  \u003ca class=\"heading-link\" href=\"#ii-unique-perspectives-and-industry-foresight\"\u003e\n    \u003ci class=\"fa-solid fa-link\" aria-hidden=\"true\" title=\"Link to heading\"\u003e\u003c/i\u003e\n    \u003cspan class=\"sr-only\"\u003eLink to heading\u003c/span\u003e\n  \u003c/a\u003e\n\u003c/h2\u003e\n\u003ctable\u003e\n  \u003cthead\u003e\n      \u003ctr\u003e\n          \u003cth style=\"text-align: left\"\u003ePerspective\u003c/th\u003e\n          \u003cth style=\"text-align: left\"\u003eSource\u003c/th\u003e\n          \u003cth style=\"text-align: left\"\u003eCore Insight\u003c/th\u003e\n      \u003c/tr\u003e\n  \u003c/thead\u003e\n  \u003ctbody\u003e\n      \u003ctr\u003e\n          \u003ctd style=\"text-align: left\"\u003e\u003cstrong\u003eAgent is the Future OS\u003c/strong\u003e\u003c/td\u003e\n          \u003ctd style=\"text-align: left\"\u003e@dotey\u003c/td\u003e\n          \u003ctd style=\"text-align: left\"\u003e\u0026ldquo;General-purpose Agents are the operating systems of the future.\u0026rdquo; Apps will diverge into: elimination (replaced by native Agent capabilities), integration (becoming Agent plugins), and enhancement (Agents augmenting original functions).\u003c/td\u003e\n      \u003c/tr\u003e\n      \u003ctr\u003e\n          \u003ctd style=\"text-align: left\"\u003e\u003cstrong\u003eAI is Integrated into the Reflex Arc\u003c/strong\u003e\u003c/td\u003e\n          \u003ctd style=\"text-align: left\"\u003e@zhixianio\u003c/td\u003e\n          \u003ctd style=\"text-align: left\"\u003eDescribes the personal experience of developing a tool using Codex\u0026rsquo;s \u003ccode\u003e/goal\u003c/code\u003e mode—\u0026ldquo;after updating the goal 5 times, the tool was ready to use,\u0026rdquo; reflecting the Agent\u0026rsquo;s autonomous iteration capability.\u003c/td\u003e\n      \u003c/tr\u003e\n      \u003ctr\u003e\n          \u003ctd style=\"text-align: left\"\u003e\u003cstrong\u003eSmall Models Fear Garbage Data, Large Models Don\u0026rsquo;t\u003c/strong\u003e\u003c/td\u003e\n          \u003ctd style=\"text-align: left\"\u003e@vista8 paraphrasing Stanford research\u003c/td\u003e\n          \u003ctd style=\"text-align: left\"\u003eOn a 15M small model, filtered data performs better across the board, but at the 330M/1B scale, \u003cstrong\u003eunfiltered data surpasses it after sufficient training\u003c/strong\u003e. Large models have enough parameters to isolate garbage from useful information.\u003c/td\u003e\n      \u003c/tr\u003e\n      \u003ctr\u003e\n          \u003ctd style=\"text-align: left\"\u003e\u003cstrong\u003eDiffusion Models as the Frontend Layer of HCI\u003c/strong\u003e\u003c/td\u003e\n          \u003ctd style=\"text-align: left\"\u003e@vista8 citing an xAI interview\u003c/td\u003e\n          \u003ctd style=\"text-align: left\"\u003e\u0026ldquo;All interfaces the user sees will be rendered in real-time by generative models\u0026rdquo;—humanity\u0026rsquo;s largest input bandwidth is vision, and the output is speech. The most natural interaction is \u0026ldquo;speaking + receiving a generative visual response.\u0026rdquo;\u003c/td\u003e\n      \u003c/tr\u003e\n      \u003ctr\u003e\n          \u003ctd style=\"text-align: left\"\u003e\u003cstrong\u003eTesting is the New Moat\u003c/strong\u003e\u003c/td\u003e\n          \u003ctd style=\"text-align: left\"\u003e@ruanyf\u003c/td\u003e\n          \u003ctd style=\"text-align: left\"\u003eResponding to a Cloudflare engineer who replicated Next.js with AI ($1100 in token fees): the moat of code is disappearing, and \u003cstrong\u003etest cases\u003c/strong\u003e are becoming the key to preventing replication.\u003c/td\u003e\n      \u003c/tr\u003e\n      \u003ctr\u003e\n          \u003ctd style=\"text-align: left\"\u003e\u003cstrong\u003eAI Programming is More Expensive Than Human Programmers\u003c/strong\u003e\u003c/td\u003e\n          \u003ctd style=\"text-align: left\"\u003e@ruanyf\u003c/td\u003e\n          \u003ctd style=\"text-align: left\"\u003eThe founder of OpenClaw consumes 603 billion tokens per month (valued at $1.3 million). Even after switching to a cheaper model, the annual cost is still 2-3 million RMB.\u003c/td\u003e\n      \u003c/tr\u003e\n      \u003ctr\u003e\n          \u003ctd style=\"text-align: left\"\u003e\u003cstrong\u003eThe Next Form of Skillsets\u003c/strong\u003e\u003c/td\u003e\n          \u003ctd style=\"text-align: left\"\u003e@lijigang\u003c/td\u003e\n          \u003ctd style=\"text-align: left\"\u003eQuestions whether plugins are the best encapsulation for Skills, proposing the \u0026ldquo;browser extension mechanism\u0026rdquo; as a possible answer.\u003c/td\u003e\n      \u003c/tr\u003e\n      \u003ctr\u003e\n          \u003ctd style=\"text-align: left\"\u003e\u003cstrong\u003eShadow Book Reading Method\u003c/strong\u003e\u003c/td\u003e\n          \u003ctd style=\"text-align: left\"\u003e@lijigang\u003c/td\u003e\n          \u003ctd style=\"text-align: left\"\u003eA reading method unique to the AI era: when reading any assertion, immediately call upon an AI to analyze opposing schools of thought, skipped premises, intellectual lineage, reasoning boundaries, etc.—\u0026ldquo;you are reading this one book plus all of its invisible counterparts.\u0026rdquo;\u003c/td\u003e\n      \u003c/tr\u003e\n  \u003c/tbody\u003e\n\u003c/table\u003e\n\u003chr\u003e\n\u003ch2 id=\"iii-recommended-tools-and-resources\"\u003e\n  III. Recommended Tools and Resources\n  \u003ca class=\"heading-link\" href=\"#iii-recommended-tools-and-resources\"\u003e\n    \u003ci class=\"fa-solid fa-link\" aria-hidden=\"true\" title=\"Link to heading\"\u003e\u003c/i\u003e\n    \u003cspan class=\"sr-only\"\u003eLink to heading\u003c/span\u003e\n  \u003c/a\u003e\n\u003c/h2\u003e\n\u003ch3 id=\"developer-tools\"\u003e\n  \u003cstrong\u003eDeveloper Tools\u003c/strong\u003e\n  \u003ca class=\"heading-link\" href=\"#developer-tools\"\u003e\n    \u003ci class=\"fa-solid fa-link\" aria-hidden=\"true\" title=\"Link to heading\"\u003e\u003c/i\u003e\n    \u003cspan class=\"sr-only\"\u003eLink to heading\u003c/span\u003e\n  \u003c/a\u003e\n\u003c/h3\u003e\n\u003ctable\u003e\n  \u003cthead\u003e\n      \u003ctr\u003e\n          \u003cth style=\"text-align: left\"\u003eTool\u003c/th\u003e\n          \u003cth style=\"text-align: left\"\u003eRecommended by\u003c/th\u003e\n          \u003cth style=\"text-align: left\"\u003ePurpose\u003c/th\u003e\n      \u003c/tr\u003e\n  \u003c/thead\u003e\n  \u003ctbody\u003e\n      \u003ctr\u003e\n          \u003ctd style=\"text-align: left\"\u003e\u003cstrong\u003eOwlia Nest\u003c/strong\u003e\u003c/td\u003e\n          \u003ctd style=\"text-align: left\"\u003e@zhixianio\u003c/td\u003e\n          \u003ctd style=\"text-align: left\"\u003eA file browsing website deployed on a PA machine, supporting Tailscale intranet access, online Markdown editing, and PWA. It solves the pain point of accessing documents produced on a remote PA.\u003c/td\u003e\n      \u003c/tr\u003e\n      \u003ctr\u003e\n          \u003ctd style=\"text-align: left\"\u003e\u003cstrong\u003eoMLX v0.4.0\u003c/strong\u003e\u003c/td\u003e\n          \u003ctd style=\"text-align: left\"\u003e@zhixianio (RT @jundotkim)\u003c/td\u003e\n          \u003ctd style=\"text-align: left\"\u003eA native Swift framework for running on-device models on macOS, with support for Native MTP.\u003c/td\u003e\n      \u003c/tr\u003e\n      \u003ctr\u003e\n          \u003ctd style=\"text-align: left\"\u003e\u003cstrong\u003eCodex-reset-watchdog\u003c/strong\u003e\u003c/td\u003e\n          \u003ctd style=\"text-align: left\"\u003e@vista8\u003c/td\u003e\n          \u003ctd style=\"text-align: left\"\u003eMonitors Codex reset messages and automatically switches to the fast model to save usage.\u003c/td\u003e\n      \u003c/tr\u003e\n      \u003ctr\u003e\n          \u003ctd style=\"text-align: left\"\u003e\u003cstrong\u003ebaoyu-image-gen Skill\u003c/strong\u003e\u003c/td\u003e\n          \u003ctd style=\"text-align: left\"\u003e@dotey\u003c/td\u003e\n          \u003ctd style=\"text-align: left\"\u003eCalls the Codex CLI within Claude Code/Hermes Agent to generate images.\u003c/td\u003e\n      \u003c/tr\u003e\n      \u003ctr\u003e\n          \u003ctd style=\"text-align: left\"\u003e\u003cstrong\u003ecc-switch\u003c/strong\u003e\u003c/td\u003e\n          \u003ctd style=\"text-align: left\"\u003e@Pluvio9yte\u003c/td\u003e\n          \u003ctd style=\"text-align: left\"\u003eA tool for switching between multiple Claude Code providers (Note: Clean up leftover \u003ccode\u003e~/.cc-switch/cc-switch.db\u003c/code\u003e configurations)\u003c/td\u003e\n      \u003c/tr\u003e\n      \u003ctr\u003e\n          \u003ctd style=\"text-align: left\"\u003e\u003cstrong\u003eHermes Desktop\u003c/strong\u003e\u003c/td\u003e\n          \u003ctd style=\"text-align: left\"\u003e@dotey (RT @NousResearch)\u003c/td\u003e\n          \u003ctd style=\"text-align: left\"\u003eNative GUI client for the Hermes Agent\u003c/td\u003e\n      \u003c/tr\u003e\n  \u003c/tbody\u003e\n\u003c/table\u003e\n\u003ch3 id=\"platforms-and-services\"\u003e\n  \u003cstrong\u003ePlatforms and Services\u003c/strong\u003e\n  \u003ca class=\"heading-link\" href=\"#platforms-and-services\"\u003e\n    \u003ci class=\"fa-solid fa-link\" aria-hidden=\"true\" title=\"Link to heading\"\u003e\u003c/i\u003e\n    \u003cspan class=\"sr-only\"\u003eLink to heading\u003c/span\u003e\n  \u003c/a\u003e\n\u003c/h3\u003e\n\u003ctable\u003e\n  \u003cthead\u003e\n      \u003ctr\u003e\n          \u003cth style=\"text-align: left\"\u003eService\u003c/th\u003e\n          \u003cth style=\"text-align: left\"\u003eRecommended by\u003c/th\u003e\n          \u003cth style=\"text-align: left\"\u003eDescription\u003c/th\u003e\n      \u003c/tr\u003e\n  \u003c/thead\u003e\n  \u003ctbody\u003e\n      \u003ctr\u003e\n          \u003ctd style=\"text-align: left\"\u003e\u003cstrong\u003eCoze 3.0\u003c/strong\u003e\u003c/td\u003e\n          \u003ctd style=\"text-align: left\"\u003e@Pluvio9yte\u003c/td\u003e\n          \u003ctd style=\"text-align: left\"\u003e\u0026ldquo;Leaps ahead of domestic competition,\u0026rdquo; with built-in Claude Code and Codex, and can connect to local Agents.\u003c/td\u003e\n      \u003c/tr\u003e\n      \u003ctr\u003e\n          \u003ctd style=\"text-align: left\"\u003e\u003cstrong\u003eFeishu Open Platform CLI\u003c/strong\u003e\u003c/td\u003e\n          \u003ctd style=\"text-align: left\"\u003e@ruanyf\u003c/td\u003e\n          \u003ctd style=\"text-align: left\"\u003eThe most feature-complete Agent integration solution among domestic office platforms, gaining over 10,000 GitHub stars in 40 days.\u003c/td\u003e\n      \u003c/tr\u003e\n      \u003ctr\u003e\n          \u003ctd style=\"text-align: left\"\u003e\u003cstrong\u003eGiffgaff\u003c/strong\u003e\u003c/td\u003e\n          \u003ctd style=\"text-align: left\"\u003e@AI_Jasonyu\u003c/td\u003e\n          \u003ctd style=\"text-align: left\"\u003eA UK SIM card, essential infrastructure for going global, with activation/usage tutorials provided.\u003c/td\u003e\n      \u003c/tr\u003e\n  \u003c/tbody\u003e\n\u003c/table\u003e\n\u003ch3 id=\"learning-resources\"\u003e\n  \u003cstrong\u003eLearning Resources\u003c/strong\u003e\n  \u003ca class=\"heading-link\" href=\"#learning-resources\"\u003e\n    \u003ci class=\"fa-solid fa-link\" aria-hidden=\"true\" title=\"Link to heading\"\u003e\u003c/i\u003e\n    \u003cspan class=\"sr-only\"\u003eLink to heading\u003c/span\u003e\n  \u003c/a\u003e\n\u003c/h3\u003e\n\u003ctable\u003e\n  \u003cthead\u003e\n      \u003ctr\u003e\n          \u003cth style=\"text-align: left\"\u003eResource\u003c/th\u003e\n          \u003cth style=\"text-align: left\"\u003eRecommended by\u003c/th\u003e\n          \u003cth style=\"text-align: left\"\u003eContent\u003c/th\u003e\n      \u003c/tr\u003e\n  \u003c/thead\u003e\n  \u003ctbody\u003e\n      \u003ctr\u003e\n          \u003ctd style=\"text-align: left\"\u003e\u003cstrong\u003eMaster 97% of Codex\u0026rsquo;s features in 30 minutes\u003c/strong\u003e\u003c/td\u003e\n          \u003ctd style=\"text-align: left\"\u003e@AI_Jasonyu (RT @servasyy_ai)\u003c/td\u003e\n          \u003ctd style=\"text-align: left\"\u003eA complete tutorial.\u003c/td\u003e\n      \u003c/tr\u003e\n      \u003ctr\u003e\n          \u003ctd style=\"text-align: left\"\u003e\u003cstrong\u003eClaude Code Workflow Breakdown\u003c/strong\u003e\u003c/td\u003e\n          \u003ctd style=\"text-align: left\"\u003e@Pluvio9yte (RT @servasyy_ai)\u003c/td\u003e\n          \u003ctd style=\"text-align: left\"\u003eAn in-depth analysis, the result of days of hard work and nearly 30 revisions.\u003c/td\u003e\n      \u003c/tr\u003e\n      \u003ctr\u003e\n          \u003ctd style=\"text-align: left\"\u003e\u003cstrong\u003eHarness Framework Tutorial\u003c/strong\u003e\u003c/td\u003e\n          \u003ctd style=\"text-align: left\"\u003e@Pluvio9yte (RT @bozhou_ai)\u003c/td\u003e\n          \u003ctd style=\"text-align: left\"\u003eLearn Harness by building a Claude Code CLI from scratch, with hands-on validation from simple to complex scenarios.\u003c/td\u003e\n      \u003c/tr\u003e\n      \u003ctr\u003e\n          \u003ctd style=\"text-align: left\"\u003e\u003cstrong\u003eStanford\u0026rsquo;s \u0026ldquo;Unfiltered Data\u0026rdquo; Paper\u003c/strong\u003e\u003c/td\u003e\n          \u003ctd style=\"text-align: left\"\u003e@vista8\u003c/td\u003e\n          \u003ctd style=\"text-align: left\"\u003eA paradigm-shifting discovery in large model training data strategies.\u003c/td\u003e\n      \u003c/tr\u003e\n      \u003ctr\u003e\n          \u003ctd style=\"text-align: left\"\u003e\u003cstrong\u003e\u0026ldquo;A Thousand Brains: A New Theory of Intelligence\u0026rdquo;\u003c/strong\u003e\u003c/td\u003e\n          \u003ctd style=\"text-align: left\"\u003e@lijigang\u003c/td\u003e\n          \u003ctd style=\"text-align: left\"\u003e\u0026ldquo;The most stunning theory of the last six months,\u0026rdquo; highlighting the inspiration of reference frame theory for cognitive AI.\u003c/td\u003e\n      \u003c/tr\u003e\n  \u003c/tbody\u003e\n\u003c/table\u003e\n\u003chr\u003e\n\u003ch2 id=\"iv-noteworthy-signals\"\u003e\n  IV. Noteworthy Signals\n  \u003ca class=\"heading-link\" href=\"#iv-noteworthy-signals\"\u003e\n    \u003ci class=\"fa-solid fa-link\" aria-hidden=\"true\" title=\"Link to heading\"\u003e\u003c/i\u003e\n    \u003cspan class=\"sr-only\"\u003eLink to heading\u003c/span\u003e\n  \u003c/a\u003e\n\u003c/h2\u003e\n\u003cul\u003e\n\u003cli\u003e\u003cstrong\u003eDiverging product paths of OpenAI and Anthropic\u003c/strong\u003e: Codex\u0026rsquo;s Sites feature directly competes with and surpasses Claude Design, showcasing OpenAI\u0026rsquo;s aggressive strategy in \u0026ldquo;end-to-end delivery.\u0026rdquo;\u003c/li\u003e\n\u003cli\u003e\u003cstrong\u003eOn-device models reaching a commercial tipping point\u003c/strong\u003e: Experience from @zhixianio shows that 35B-scale models are now production-ready on consumer-grade hardware, potentially accelerating the explosion of edge AI applications.\u003c/li\u003e\n\u003cli\u003e\u003cstrong\u003eDomestic regulatory challenges\u003c/strong\u003e: @Pluvio9yte bluntly stated, \u0026ldquo;Compliance for AI products in China is incredibly difficult\u0026hellip; before you even earn any money, you have to go through the painful process of applying for a pile of certificates.\u0026rdquo;\u003c/li\u003e\n\u003cli\u003e\u003cstrong\u003eRestructuring of Tokenomics\u003c/strong\u003e: The \u0026ldquo;unlimited usage costs\u0026rdquo; revealed by @ruanyf stand in stark contrast to @gefei55\u0026rsquo;s observation that \u0026ldquo;people are buying the $150/month subscription,\u0026rdquo; signaling an extreme divergence in pricing strategies in the age of Agents.\u003c/li\u003e\n\u003c/ul\u003e\n\u003ch2 id=\"-appendix-todays-watch-list-source-updates\"\u003e\n  📚 Appendix: Today\u0026rsquo;s Watch List Source Updates\n  \u003ca class=\"heading-link\" href=\"#-appendix-todays-watch-list-source-updates\"\u003e\n    \u003ci class=\"fa-solid fa-link\" aria-hidden=\"true\" title=\"Link to heading\"\u003e\u003c/i\u003e\n    \u003cspan class=\"sr-only\"\u003eLink to heading\u003c/span\u003e\n  \u003c/a\u003e\n\u003c/h2\u003e\n\u003cblockquote\u003e\n\u003cp\u003eTimeframe: Last 3 days; 16 sources covered; 9 updates in total.\u003c/p\u003e\n\u003c/blockquote\u003e\n\u003ch3 id=\"y-combinator-podcast-b_introsearch\"\u003e\n  Y Combinator Podcast (B_intro+search)\n  \u003ca class=\"heading-link\" href=\"#y-combinator-podcast-b_introsearch\"\u003e\n    \u003ci class=\"fa-solid fa-link\" aria-hidden=\"true\" title=\"Link to heading\"\u003e\u003c/i\u003e\n    \u003cspan class=\"sr-only\"\u003eLink to heading\u003c/span\u003e\n  \u003c/a\u003e\n\u003c/h3\u003e\n\u003cul\u003e\n\u003cli\u003e\u003cstrong\u003e\u003ca href=\"https://podcasters.spotify.com/pod/show/ycombinator/episodes/How-to-Build-an-AI-Native-Services-Company-e3k8dj5\"  class=\"external-link\" target=\"_blank\" rel=\"noopener\"\u003eHow to Build an AI-Native Services Company\u003c/a\u003e\u003c/strong\u003e\n\u003cul\u003e\n\u003cli\u003ePublished: 2026-06-04 00:00 Beijing Time\u003c/li\u003e\n\u003cli\u003eSummary: - You\u0026rsquo;ve probably heard of OpenClaw (formerly Clawdbot/Moltbot).\n\u003cul\u003e\n\u003cli\u003eThe sensational open-source AI assistant that runs on your own device, connects with the messaging apps you already use, and goes beyond chat to actually do things like manage your email, calendar, files, workflows, and more.\u003c/li\u003e\n\u003cli\u003eNow, meet the person behind it.\u003c/li\u003e\n\u003cli\u003eYC\u0026rsquo;s Raphael Schaad sits down with OpenClaw founder Peter Steinberger to talk about the \u0026ldquo;aha\u0026rdquo; moment behind the viral personal AI agent, why local-first agents could replace many of today\u0026rsquo;s apps, and how personal agents will reshape the future of software.\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003cli\u003eEN Highlights:\n\u003cul\u003e\n\u003cli\u003eSome of the biggest companies of the next decade won\u0026rsquo;t be software businesses, they\u0026rsquo;ll be services companies like insurance carriers, law firms, and tax practic…\u003c/li\u003e\n\u003cli\u003eIn this episode of Startup School, YC Visiting Partner Charlie Warren walks through the playbook for building AI native services companies, covering how to pick…\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003c/ul\u003e\n\u003ch3 id=\"all-in-podcast-a_full\"\u003e\n  All-In Podcast (A_full)\n  \u003ca class=\"heading-link\" href=\"#all-in-podcast-a_full\"\u003e\n    \u003ci class=\"fa-solid fa-link\" aria-hidden=\"true\" title=\"Link to heading\"\u003e\u003c/i\u003e\n    \u003cspan class=\"sr-only\"\u003eLink to heading\u003c/span\u003e\n  \u003c/a\u003e\n\u003c/h3\u003e\n\u003cul\u003e\n\u003cli\u003e\u003cstrong\u003e\u003ca href=\"https://allinchamathjason.libsyn.com/bill-ackman-investment-strategy-what-the-market-is-missing-how-ai-breaks-businesses\"  class=\"external-link\" target=\"_blank\" rel=\"noopener\"\u003eBill Ackman: Investment Strategy, What the Market is Missing, How AI Breaks Businesses\u003c/a\u003e\u003c/strong\u003e\n\u003cul\u003e\n\u003cli\u003ePublished: 2026-06-03 11:06 Beijing Time\u003c/li\u003e\n\u003cli\u003eSummary: - EY - Agentic AI is introducing new investment rules.\n\u003cul\u003e\n\u003cli\u003eAs AI shifts to a consumption-based model, EY is linking spending to business value.\u003c/li\u003e\n\u003cli\u003eNYSE - Thanks to our partner the New York Stock Exchange - a modern marketplace and exchange dedicated to building the future.\u003c/li\u003e\n\u003cli\u003ePlaud, our official wearable AI note-taking partner at the All-In Liquidity Summit, captured every insight.\u003c/li\u003e\n\u003cli\u003eBill Ackman: Investment strategy, what the market is missing, and how AI is breaking businesses.\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003cli\u003eEN Highlights:\n\u003cul\u003e\n\u003cli\u003e(0:00) Bill Ackman joins the show\u003c/li\u003e\n\u003cli\u003e(0:30) Evolving investment philosophy: What\u0026rsquo;s changed over 20 years\u003c/li\u003e\n\u003cli\u003e(4:40) AI: Greatest time to build a business, and a major threat to portfolios\u003c/li\u003e\n\u003cli\u003e(7:50) Predicting market moves, the \u0026ldquo;rubber band effect\u0026rdquo;\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003c/ul\u003e\n\u003ch3 id=\"stratechery-by-ben-thompson-a_full\"\u003e\n  Stratechery by Ben Thompson (A_full)\n  \u003ca class=\"heading-link\" href=\"#stratechery-by-ben-thompson-a_full\"\u003e\n    \u003ci class=\"fa-solid fa-link\" aria-hidden=\"true\" title=\"Link to heading\"\u003e\u003c/i\u003e\n    \u003cspan class=\"sr-only\"\u003eLink to heading\u003c/span\u003e\n  \u003c/a\u003e\n\u003c/h3\u003e\n\u003cul\u003e\n\u003cli\u003e\u003cstrong\u003e\u003ca href=\"https://stratechery.com/2026/the-nvidia-ai-pc-project-solara-microsoft-ai/\"  class=\"external-link\" target=\"_blank\" rel=\"noopener\"\u003eThe Nvidia AI PC, Project Solara, Microsoft AI\u003c/a\u003e\u003c/strong\u003e\n\u003cul\u003e\n\u003cli\u003ePublished: 2026-06-03 18:00 Beijing Time\u003c/li\u003e\n\u003cli\u003eSummary: - I don\u0026rsquo;t usually reveal my interview topics in advance, but I\u0026rsquo;m making an exception this week given the subject and the update below.\n\u003cul\u003e\n\u003cli\u003eHowever, in this case, I came to the following conclusions during the keynote and before the interview, so for that reason (and some logistical reasons), I want to lay them out first (before you see my questions), and then follow up with Nadella\u0026rsquo;s take on them (and many other topics).\u003c/li\u003e\n\u003cli\u003eWith that said, here is the update:\n-\u0026gt; Nvidia became the world\u0026rsquo;s most valuable company by dominating the data center AI chip market.\u003c/li\u003e\n\u003cli\u003eNow, the company is extending its strength to chips that will serve as the main processors for personal computers, entering a field long dominated by Intel, Advanced Micro Devices, Qualcomm, and Apple.\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003cli\u003eEN Highlights:\n\u003cul\u003e\n\u003cli\u003eListen to this post:\u003c/li\u003e\n\u003cli\u003eGood morning,\u003c/li\u003e\n\u003cli\u003eI don’t normally give away my interview subjects ahead of time, but I’m going to make an exception this week given the subject and the below Update\u003c/li\u003e\n\u003cli\u003eI am writing this in San Francisco where I interviewed Microsoft CEO Satya Nadella after his Build developer conference keynote ; normally I would want to publi…\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003c/ul\u003e\n\u003ch3 id=\"openai-blog-a_full\"\u003e\n  OpenAI Blog (A_full)\n  \u003ca class=\"heading-link\" href=\"#openai-blog-a_full\"\u003e\n    \u003ci class=\"fa-solid fa-link\" aria-hidden=\"true\" title=\"Link to heading\"\u003e\u003c/i\u003e\n    \u003cspan class=\"sr-only\"\u003eLink to heading\u003c/span\u003e\n  \u003c/a\u003e\n\u003c/h3\u003e\n\u003cul\u003e\n\u003cli\u003e\n\u003cp\u003e\u003cstrong\u003e\u003ca href=\"https://openai.com/index/introducing-new-capabilities-to-gpt-rosalind\"  class=\"external-link\" target=\"_blank\" rel=\"noopener\"\u003eIntroducing new capabilities to GPT-Rosalind\u003c/a\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cul\u003e\n\u003cli\u003ePublished: 2026-06-03 21:15 Beijing Time\u003c/li\u003e\n\u003cli\u003eSummary: - It combines the agentic coding and tool-use capabilities of GPT-5.5 with more powerful model intelligence in core drug discovery domains like medicinal chemistry and genomics, while improving performance across a broader range of life sciences analysis, design, and experimental workflows.\n\u003cul\u003e\n\u003cli\u003eAdvances in life sciences depend on the synthesis of data and evidence across scales and modalities: molecules, genes, pathways, and living systems.\u003c/li\u003e\n\u003cli\u003eIn our evaluations, the updated GPT-Rosalind shows broad performance improvements on expert biologist research tasks, complex medicinal chemistry queries, quantitative biology, and wet lab troubleshooting.\u003c/li\u003e\n\u003cli\u003eGPT-Rosalind is now available in research preview to eligible organizations worldwide through our trusted access deployment structure.\u003c/li\u003e\n\u003cli\u003e\n\u003ch2 id=\"improving-performance-on-scientifically-valuable-tasks\"\u003e\n  Improving performance on scientifically valuable tasks.\n  \u003ca class=\"heading-link\" href=\"#improving-performance-on-scientifically-valuable-tasks\"\u003e\n    \u003ci class=\"fa-solid fa-link\" aria-hidden=\"true\" title=\"Link to heading\"\u003e\u003c/i\u003e\n    \u003cspan class=\"sr-only\"\u003eLink to heading\u003c/span\u003e\n  \u003c/a\u003e\n\u003c/h2\u003e\n\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003cli\u003eEN Highlights:\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003cli\u003e\n\u003cp\u003eGPT-Rosalind advances life sciences research with enhanced biological reasoning, medicinal chemistry expertise, genomics analysis, and experimental workflow cap…\u003c/p\u003e\n\u003c/li\u003e\n\u003cli\u003e\n\u003cp\u003e\u003cstrong\u003e\u003ca href=\"https://openai.com/index/wasmer\"  class=\"external-link\" target=\"_blank\" rel=\"noopener\"\u003eHow Wasmer used Codex to build a Node.js runtime for the edge\u003c/a\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cul\u003e\n\u003cli\u003eRelease Time: 2026-06-03 20:00 Beijing Time\u003c/li\u003e\n\u003cli\u003eSummary: - Wasmer\u0026rsquo;s engineers made a breakthrough this year: they figured out how to run Node.js workloads inside a WebAssembly sandbox, enabling developers to run JavaScript applications, MCPs, and proxies without Docker.\n\u003cul\u003e\n\u003cli\u003eWithout Codex, this work would have taken a year, but with Codex, it took two weeks.\u003c/li\u003e\n\u003cli\u003eNow, they are the first cloud host to offer full Node.js at the edge layer.\n-\u0026gt; \u0026ldquo;We\u0026rsquo;re actually moving away from the IDE itself.\u003c/li\u003e\n\u003cli\u003eWe\u0026rsquo;re not touching the code that much, we\u0026rsquo;re just guiding it where we want it to go.\u0026rdquo;\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003cli\u003eEN Highlights:\n\u003cul\u003e\n\u003cli\u003eSee how Wasmer used Codex with GPT-5.5 to build a Node.js runtime for the edge, accelerating development 10x to 20x and shipping in weeks instead of months.\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003cli\u003e\n\u003cp\u003e\u003cstrong\u003e\u003ca href=\"https://openai.com/index/public-policy-agenda\"  class=\"external-link\" target=\"_blank\" rel=\"noopener\"\u003eOpenAI public policy agenda\u003c/a\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cul\u003e\n\u003cli\u003eRelease Time: 2026-06-03 18:00 Beijing Time\u003c/li\u003e\n\u003cli\u003eSummary: - OpenAI outlines its public policy agenda for artificial intelligence, including safety, youth protection, workforce transition, and global standards to ensure AI benefits society.\n\u003cul\u003e\n\u003cli\u003eThis post from the OpenAI blog explains how OpenAI\u0026rsquo;s public policy agenda is shaping the broader AI and infrastructure landscape.\u003c/li\u003e\n\u003cli\u003eIt also provides practical implications for founders, operators, and investors following OpenAI\u0026rsquo;s public policy agenda.\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003cli\u003eEN Highlights:\n\u003cul\u003e\n\u003cli\u003eOpenAI outlines its public policy agenda for AI, including safety, youth protection, workforce transition, and global standards to ensure AI benefits society.\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003cli\u003e\n\u003cp\u003e\u003cstrong\u003e\u003ca href=\"https://openai.com/index/frontier-safety-blueprint\"  class=\"external-link\" target=\"_blank\" rel=\"noopener\"\u003eA blueprint for democratic governance of frontier AI\u003c/a\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cul\u003e\n\u003cli\u003eRelease Time: 2026-06-03 18:00 Beijing Time\u003c/li\u003e\n\u003cli\u003eSummary: - Governance of frontier AI, proposing a federal framework for safety, resilience, and national security.\n\u003cul\u003e\n\u003cli\u003eOpenAI outlines a blueprint for U.S. governance of frontier AI, proposing a federal framework for safety, resilience, and national security.\u003c/li\u003e\n\u003cli\u003eA blueprint for the democratic governance of frontier AI.\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003cli\u003eEN Highlights:\n\u003cul\u003e\n\u003cli\u003eOpenAI outlines a blueprint for U.S\u003c/li\u003e\n\u003cli\u003egovernance of frontier AI, proposing a federal framework for safety, resilience, and national security.\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003c/ul\u003e\n\u003ch3 id=\"two-minute-papers-b_introsearch\"\u003e\n  Two Minute Papers (B_intro+search)\n  \u003ca class=\"heading-link\" href=\"#two-minute-papers-b_introsearch\"\u003e\n    \u003ci class=\"fa-solid fa-link\" aria-hidden=\"true\" title=\"Link to heading\"\u003e\u003c/i\u003e\n    \u003cspan class=\"sr-only\"\u003eLink to heading\u003c/span\u003e\n  \u003c/a\u003e\n\u003c/h3\u003e\n\u003cul\u003e\n\u003cli\u003e\n\u003cp\u003e\u003cstrong\u003e\u003ca href=\"https://www.youtube.com/shorts/iz7elwbqZVI\"  class=\"external-link\" target=\"_blank\" rel=\"noopener\"\u003eMeet the AI \u0026ldquo;Co-Scientist\u0026rdquo; Changing Everything 🤖🧪 #ai\u003c/a\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cul\u003e\n\u003cli\u003eRelease Time: 2026-06-04 01:00 Beijing Time\u003c/li\u003e\n\u003cli\u003eSummary: - Meet the AI \u0026ldquo;Co-Scientist\u0026rdquo; Changing Everything 🤖🧪 #ai.\n\u003cul\u003e\n\u003cli\u003eThis update provides builders with a quick background on Meet the AI \u0026ldquo;Co-Scientist\u0026rdquo; Changing Everything 🤖🧪 #ai and why it\u0026rsquo;s worth bookmarking.\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003cli\u003eEN Highlights:\n\u003cul\u003e\n\u003cli\u003eMeet the AI \u0026ldquo;Co-Scientist\u0026rdquo; Changing Everything 🤖🧪 #ai\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003cli\u003e\n\u003cp\u003e\u003cstrong\u003e\u003ca href=\"https://www.youtube.com/watch?v=ypL7kUiw_LM\"  class=\"external-link\" target=\"_blank\" rel=\"noopener\"\u003eClaude Opus 4.8: Lying Machine No More?\u003c/a\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cul\u003e\n\u003cli\u003eRelease Time: 2026-06-03 21:49 Beijing Time\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003cli\u003e\n\u003cp\u003eSummary:\u003c/p\u003e\n\u003cul\u003e\n\u003cli\u003e❤️ Check out Lambda here and sign up for their GPU Cloud:\u003c/li\u003e\n\u003cli\u003eAdam Bridges, Benji Rabhan, B Shang, Cameron Navor, Charles Ian Norman Venn, Christian Ahlin, Eric T, Fred R, Gordon Child, Juan Benet, Michael Tedder, Owen Skarpness, Richard Sundvall, Ryan Stankye, Shawn Becker, Steef, Taras Bobrovytsky, Tazaur Sagenclaw, Tybie Fitzhugh, Ueli Gallizzi.\u003c/li\u003e\n\u003cli\u003eClaude Opus 4.8: Are lying machines no more?\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003c/ul\u003e\n",
  "wordCount": 3721,
  "readingTime": 18,
  "tableOfContents": "\u003cnav id=\"TableOfContents\"\u003e\n  \u003cul\u003e\n    \u003cli\u003e\u003ca href=\"#-in-depth-guide-to-this-issues-watch-list\"\u003e📖 In-depth Guide to This Issue\u0026rsquo;s Watch List\u003c/a\u003e\u003c/li\u003e\n    \u003cli\u003e\u003ca href=\"#-trending-ai-news-on-x\"\u003e🌐 Trending AI News on X\u003c/a\u003e\n      \u003cul\u003e\n        \u003cli\u003e\u003ca href=\"#topic-1-openclaw-202661-brings-native-windows-support-and-self-improving-agents\"\u003eTopic 1: OpenClaw 2026.6.1 Brings Native Windows Support and Self-Improving Agents\u003c/a\u003e\u003c/li\u003e\n        \u003cli\u003e\u003ca href=\"#topic-2-anthropic-engineers-share-free-claude-code-mastery-talks\"\u003eTopic 2: Anthropic Engineers Share Free Claude Code Mastery Talks\u003c/a\u003e\u003c/li\u003e\n        \u003cli\u003e\u003ca href=\"#topic-3-openais-its-time-to-fly-video-spotlights-codex-coding-power\"\u003eTopic 3: OpenAI\u0026rsquo;s \u0026lsquo;It\u0026rsquo;s Time to Fly\u0026rsquo; Video Spotlights Codex Coding Power\u003c/a\u003e\u003c/li\u003e\n        \u003cli\u003e\u003ca href=\"#topic-4-openai-codex-team-aims-for-ultimate-desktop-coding-app\"\u003eTopic 4: OpenAI Codex Team Aims for Ultimate Desktop Coding App\u003c/a\u003e\u003c/li\u003e\n        \u003cli\u003e\u003ca href=\"#topic-5-reve-20-launches-as-top-ai-image-generator-with-layout-control\"\u003eTopic 5: Reve 2.0 Launches as Top AI Image Generator with Layout Control\u003c/a\u003e\u003c/li\u003e\n        \u003cli\u003e\u003ca href=\"#topic-6-openais-codex-reaches-5-million-weekly-users-as-productivity-tool\"\u003eTopic 6: OpenAI\u0026rsquo;s Codex Reaches 5 Million Weekly Users as Productivity Tool\u003c/a\u003e\u003c/li\u003e\n        \u003cli\u003e\u003ca href=\"#topic-7-nvidias-jensen-huang-shares-beer-on-computex-floor-with-gigabyte-exec\"\u003eTopic 7: NVIDIA\u0026rsquo;s Jensen Huang Shares Beer on Computex Floor with Gigabyte Exec\u003c/a\u003e\u003c/li\u003e\n      \u003c/ul\u003e\n    \u003c/li\u003e\n    \u003cli\u003e\u003ca href=\"#-influencer-insights\"\u003e💡 Influencer Insights\u003c/a\u003e\u003c/li\u003e\n  \u003c/ul\u003e\n\n  \u003cul\u003e\n    \u003cli\u003e\u003ca href=\"#i-widely-followed-tech-trends-and-product-highlights\"\u003eI. Widely Followed Tech Trends and Product Highlights\u003c/a\u003e\n      \u003cul\u003e\n        \u003cli\u003e\u003ca href=\"#1-the-codex-ecosystem-continues-to-explode-becoming-the-top-choice-for-agent-development\"\u003e1. \u003cstrong\u003eThe Codex ecosystem continues to explode, becoming the top choice for Agent development\u003c/strong\u003e\u003c/a\u003e\u003c/li\u003e\n        \u003cli\u003e\u003ca href=\"#2-comparison-and-migration-between-claude-code-and-codex\"\u003e2. \u003cstrong\u003eComparison and Migration between Claude Code and Codex\u003c/strong\u003e\u003c/a\u003e\u003c/li\u003e\n        \u003cli\u003e\u003ca href=\"#3-leap-in-on-device-model-capabilities-local-deployment-becomes-a-new-trend\"\u003e3. \u003cstrong\u003eLeap in On-Device Model Capabilities, Local Deployment Becomes a New Trend\u003c/strong\u003e\u003c/a\u003e\u003c/li\u003e\n        \u003cli\u003e\u003ca href=\"#4-agent-memory-solutions-and-workflow-optimization\"\u003e4. \u003cstrong\u003eAgent Memory Solutions and Workflow Optimization\u003c/strong\u003e\u003c/a\u003e\u003c/li\u003e\n      \u003c/ul\u003e\n    \u003c/li\u003e\n    \u003cli\u003e\u003ca href=\"#ii-unique-perspectives-and-industry-foresight\"\u003eII. Unique Perspectives and Industry Foresight\u003c/a\u003e\u003c/li\u003e\n    \u003cli\u003e\u003ca href=\"#iii-recommended-tools-and-resources\"\u003eIII. Recommended Tools and Resources\u003c/a\u003e\n      \u003cul\u003e\n        \u003cli\u003e\u003ca href=\"#developer-tools\"\u003e\u003cstrong\u003eDeveloper Tools\u003c/strong\u003e\u003c/a\u003e\u003c/li\u003e\n        \u003cli\u003e\u003ca href=\"#platforms-and-services\"\u003e\u003cstrong\u003ePlatforms and Services\u003c/strong\u003e\u003c/a\u003e\u003c/li\u003e\n        \u003cli\u003e\u003ca href=\"#learning-resources\"\u003e\u003cstrong\u003eLearning Resources\u003c/strong\u003e\u003c/a\u003e\u003c/li\u003e\n      \u003c/ul\u003e\n    \u003c/li\u003e\n    \u003cli\u003e\u003ca href=\"#iv-noteworthy-signals\"\u003eIV. Noteworthy Signals\u003c/a\u003e\u003c/li\u003e\n    \u003cli\u003e\u003ca href=\"#-appendix-todays-watch-list-source-updates\"\u003e📚 Appendix: Today\u0026rsquo;s Watch List Source Updates\u003c/a\u003e\n      \u003cul\u003e\n        \u003cli\u003e\u003ca href=\"#y-combinator-podcast-b_introsearch\"\u003eY Combinator Podcast (B_intro+search)\u003c/a\u003e\u003c/li\u003e\n        \u003cli\u003e\u003ca href=\"#all-in-podcast-a_full\"\u003eAll-In Podcast (A_full)\u003c/a\u003e\u003c/li\u003e\n        \u003cli\u003e\u003ca href=\"#stratechery-by-ben-thompson-a_full\"\u003eStratechery by Ben Thompson (A_full)\u003c/a\u003e\u003c/li\u003e\n        \u003cli\u003e\u003ca href=\"#openai-blog-a_full\"\u003eOpenAI Blog (A_full)\u003c/a\u003e\u003c/li\u003e\n      \u003c/ul\u003e\n    \u003c/li\u003e\n    \u003cli\u003e\u003ca href=\"#improving-performance-on-scientifically-valuable-tasks\"\u003eImproving performance on scientifically valuable tasks.\u003c/a\u003e\n      \u003cul\u003e\n        \u003cli\u003e\u003ca href=\"#two-minute-papers-b_introsearch\"\u003eTwo Minute Papers (B_intro+search)\u003c/a\u003e\u003c/li\u003e\n      \u003c/ul\u003e\n    \u003c/li\u003e\n  \u003c/ul\u003e\n\u003c/nav\u003e",
  "isDraft": false
}