今日重点

• TOP PICK · 评分 100.0
  Opus 5.5 vs. GPT-6 Sol: which model won my blind taste test?
  Listen or watch on YouTube , Spotify , or Apple Podcasts I got up ear…
• TOP PICK · 评分 100.0
  Where're All The AI Chips?
  If you liked this piece, you should subscribe to my premium newslette…
• TOP PICK · 评分 100.0
  tokens too cheap to meter
  The price of using machine learning intelligence is decreasing by sev…

深度阅读

• Opus 5.5 vs. GPT-6 Sol: which model won my blind taste test?
  Lenny's Newsletter · 09-23 07:12 · 评分 100.0
  Listen or watch on YouTube , Spotify , or Apple Podcasts I got up ear…
  高优先级：可能影响 AI 产品、模型能力或产业判断。
  OpenAI Anthropic Agent
• Claude Opus 5.5, GPT-6 Sol, GPT-6 Luna, and a new price war
  Simon Willison · 未知时间 · 评分 100.0
  Yesterday was Grok 4.7 ( pelicans ) and MiMo v2.6 Flash/Pro ( more pe…
  高优先级：可能影响 AI 产品、模型能力或产业判断。
  OpenAI Anthropic Agent
• I left Claude for months. Opus 5.5 is why I'm back
  Lenny's Newsletter · 09-23 03:06 · 评分 96.2
  Listen or watch on YouTube , Spotify , or Apple Podcasts I’ve been of…
  高优先级：可能影响 AI 产品、模型能力或产业判断。
  OpenAI Anthropic Agent
• Advanced evals: How to find (and fix) hidden AI failures in your product
  Lenny's Newsletter · 09-22 20:45 · 评分 79.7
  👋 Hey there, I’m Lenny. Each week, I share deeply researched product,…
  中高优先级：具备较强信息密度或行业参考价值。
  Meta Anthropic Agent
• llm 0.36
  Simon Willison · 未知时间 · 评分 78.5
  Release: llm 0.36 New OpenAI models: gpt-6-sol for GPT-6 Sol and gpt-…
  中高优先级：具备较强信息密度或行业参考价值。
  OpenAI

HN 热门博客 2025

• Where're All The AI Chips?
  wheresyoured.at · 09-22 22:35 · 评分 100.0
  If you liked this piece, you should subscribe to my premium newslette…
  高优先级：可能影响 AI 产品、模型能力或产业判断。
  OpenAI Google Meta Anthropic
• tokens too cheap to meter
  jyn.dev · 未知时间 · 评分 100.0
  The price of using machine learning intelligence is decreasing by sev…
  高优先级：可能影响 AI 产品、模型能力或产业判断。
  OpenAI Google Anthropic Agent
• Meta’s New Muse AI Agent App Overtakes ChatGPT as Top iPhone App
  daringfireball.net · 未知时间 · 评分 100.0
  Zac Hall, writing for 9to5 Mac on Friday: Meta’s new Muse AI agent ap…
  高优先级：可能影响 AI 产品、模型能力或产业判断。
  OpenAI Google Meta Anthropic
• Unfinished Work in Package Security
  nesbitt.io · 未知时间 · 评分 87.0
  In May 2026, an Nx contributor installed a malicious dependency that…
  高优先级：可能影响 AI 产品、模型能力或产业判断。
  Meta Agent API
• Pluralistic: Bonta sold us out to Trump's oligarchs (22 Sep 2026)
  pluralistic.net · 09-22 16:19 · 评分 73.0
  Today's links Bonta sold us out to Trump's oligarchs : A "settlement"…
  中高优先级：具备较强信息密度或行业参考价值。
  Agent API
• Package Manager Threat Model, Revisited
  nesbitt.io · 未知时间 · 评分 73.0
  I’ve spent a fair chunk of the last four months pointing the package…
  中高优先级：具备较强信息密度或行业参考价值。
  Agent API
• A Great Story Capturing the Essence of Being Online in 2026
  blog.jim-nielsen.com · 09-22 03:00 · 评分 64.9
  I made a note of it , but Bryan Cantrill wrote a post about the ex-An…
  一般优先级：适合补充背景和趋势感知。
  Google Anthropic
• Cheaper LLM labelling
  entropicthoughts.com · 09-22 06:00 · 评分 63.6
  I have a small project where I needed to label commits as either “mai…
  一般优先级：适合补充背景和趋势感知。
  OpenAI
• Why Didn’t Google Build Muse?
  daringfireball.net · 未知时间 · 评分 63.2
  MG Siegler, writing at Spyglass: The number one response I’ve gotten…
  一般优先级：适合补充背景和趋势感知。
  Google Meta Agent

科技与商业

• Agent 版 Hugging Face 来了，openJiuwen 发布首个开源智能体资产平台
  Readhub · 09-23 04:50 · 评分 76.5
  openJiuwen 推出首个开源智能体资产平台，将 Agent 周边能力拆分为技能、连接器、插件、专家和专家团五类可管理、可复用、可共享…
  中高优先级：具备较强信息密度或行业参考价值。
  开源 Agent

AI 核心

• Parallel cut research time and cost in half with GPT‑6 Astra
  OpenAI Blog · 09-22 20:00 · 评分 73.3
  GPT‑6 Astra allowed Parallel’s agents to research and synthesize labo…
  中高优先级：具备较强信息密度或行业参考价值。
  Agent
• Introducing GPT-6 Sol and Luna
  OpenAI Blog · 09-23 02:00 · 评分 71.3
  Meet GPT-6 Sol and Luna, two models that bring frontier intelligence…
  中高优先级：具备较强信息密度或行业参考价值。
• Priorities and principles for effective third party assessments
  OpenAI Blog · 09-22 08:00 · 评分 67.4
  OpenAI outlines priorities and principles for rigorous, secure, and i…
  中高优先级：具备较强信息密度或行业参考价值。
  OpenAI
• Better prompt caching for GPT-6
  OpenAI Blog · 09-23 05:00 · 评分 66.4
  Learn how GPT-6 improves prompt caching with higher cache hit rates,…
  中高优先级：具备较强信息密度或行业参考价值。

🤖 AI/人工智能相关

• ChatGPT版GrokBot代码曝光！OpenAI也要为你造不下班的AI同事
  新智元 · 未知时间 · 评分 71.1
  ChatGPT版GrokBot代码曝光！OpenAI也要为你造不下班的AI同事
  中高优先级：具备较强信息密度或行业参考价值。
  OpenAI