<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:dc="http://purl.org/dc/elements/1.1/">
<channel>
  <title>AI 武林：OpenCL</title>
  <link>https://aiwulin.itsmygo.uk/tools/opencl/</link>
  <description>AI 武林收錄的內容裡，最新提到 OpenCL 的 24 筆，每筆附中文摘要。</description>
  <language>zh-TW</language>
  <lastBuildDate>Thu, 08 Oct 2026 17:22:41 GMT</lastBuildDate>
  <atom:link href="https://aiwulin.itsmygo.uk/tools/opencl/rss.xml" rel="self" type="application/rss+xml"/>
  <item>
    <title>Why Everyone is Now Getting Excited About Personal AI Agents</title>
    <link>https://aiwulin.itsmygo.uk/c/4a9e7de7b9/</link>
    <guid isPermaLink="false">aiwulin-4a9e7de7b9</guid>
    <pubDate>Sat, 19 Sep 2026 00:00:00 +0800</pubDate>
    <dc:creator>The AI Daily Brief: Artificial Intelligence News</dc:creator>
    <description>&lt;p&gt;Meta 的 Muse 個人 AI 代理在 App Store 排名突升，顯示消費者端正從商業應用轉向個人使用。影片探討了推動此變化的設計模式，如持續性、目標建立與智慧預設，並分析 OpenAI、Google DeepMind 等公司的最…&lt;/p&gt;&lt;p&gt;原站：&lt;a href=&quot;https://www.youtube.com/watch?v=NmXhFS39ad0&quot;&gt;https://www.youtube.com/watch?v=NmXhFS39ad0&lt;/a&gt;&lt;/p&gt;</description>
    <category>影片</category>
    <category>AI Agent 基礎</category>
  </item>
  <item>
    <title>ThursdAI - Sep 17 - TypeSafe's Jev is a ChatGPT moment for decisions, Pace the Frontier splits the labs &amp; more</title>
    <link>https://aiwulin.itsmygo.uk/c/56211cef7d/</link>
    <guid isPermaLink="false">aiwulin-56211cef7d</guid>
    <pubDate>Fri, 18 Sep 2026 00:00:00 +0800</pubDate>
    <dc:creator>ThursdAI</dc:creator>
    <description>&lt;p&gt;訪談探討 TypeSafe 推出的非 LLM 推理模型 Jev，強調其速度與成本優勢；同時分析 Dario Amodei 提出的「緩衝前沿」策略、各大廠對 AI 助理的競爭，以及政治人物對 AI 發展的看法。&lt;/p&gt;&lt;p&gt;原站：&lt;a href=&quot;https://sub.thursdai.news/p/typesafes-jev-changes-everything&quot;&gt;https://sub.thursdai.news/p/typesafes-jev-changes-everything&lt;/a&gt;&lt;/p&gt;</description>
    <category>Podcast</category>
    <category>研究前沿</category>
    <category>AI 工作術</category>
    <category>大型語言模型入門</category>
  </item>
  <item>
    <title>How to Build Your Own AI Agent Team From Scratch</title>
    <link>https://aiwulin.itsmygo.uk/c/fe65d4dce5/</link>
    <guid isPermaLink="false">aiwulin-fe65d4dce5</guid>
    <pubDate>Mon, 14 Sep 2026 00:00:00 +0800</pubDate>
    <dc:creator>Tech With Tim</dc:creator>
    <description>&lt;p&gt;展示如何利用 Hyper Agent 平台建立一個由四個 AI 代理組成的團隊，自動處理贊助商審核、風險評估、指令碼構思與簡報製作。觀眾可學習如何從零開始設定代理、連線 Slack 與 Google 工具，並透過任務連結讓代理協同工作。&lt;/p&gt;&lt;p&gt;原站：&lt;a href=&quot;https://www.youtube.com/watch?v=0nU4lQnnSz0&quot;&gt;https://www.youtube.com/watch?v=0nU4lQnnSz0&lt;/a&gt;&lt;/p&gt;</description>
    <category>影片</category>
    <category>AI Agent 基礎</category>
  </item>
  <item>
    <title>The Multiplayer AI Sprint: Build Your Team’s First Shared Agent</title>
    <link>https://aiwulin.itsmygo.uk/c/c324c8ccce/</link>
    <guid isPermaLink="false">aiwulin-c324c8ccce</guid>
    <pubDate>Fri, 11 Sep 2026 00:00:00 +0800</pubDate>
    <dc:creator>The AI Daily Brief: Artificial Intelligence News</dc:creator>
    <description>&lt;p&gt;探討 AI 從個人工具轉向團隊共享代理的趨勢，介紹如何建立團隊共用的 AI 工作空間。觀眾可學習如何評估並實施團隊共享代理，提升跨職能協作效率。&lt;/p&gt;&lt;p&gt;原站：&lt;a href=&quot;https://www.youtube.com/watch?v=1BqWLLkJym0&quot;&gt;https://www.youtube.com/watch?v=1BqWLLkJym0&lt;/a&gt;&lt;/p&gt;</description>
    <category>影片</category>
    <category>AI Agent 基礎</category>
  </item>
  <item>
    <title>OpenAI solves Navier-Stokes, Meta’s Muse a free AI agent that’s really good, DeepSeek V4.1 shrinks KV cache, and one doomer post causes OpenAI to consider pausing training + more AI news</title>
    <link>https://aiwulin.itsmygo.uk/c/e3423e6340/</link>
    <guid isPermaLink="false">aiwulin-e3423e6340</guid>
    <pubDate>Fri, 11 Sep 2026 00:00:00 +0800</pubDate>
    <dc:creator>ThursdAI</dc:creator>
    <description>&lt;p&gt;Podcast 涵蓋 OpenAI 宣稱 1 萬個代理解決納維 - 斯托克斯方程、DeepSeek V4.1 發布及壓縮 KV Cache、Meta 推出免費 AI 代理 Muse 與保密虛擬機器、以及一名前員工因擔憂 AI 滅世而引發的…&lt;/p&gt;&lt;p&gt;原站：&lt;a href=&quot;https://sub.thursdai.news/p/openai-solves-navier-stokes-metas&quot;&gt;https://sub.thursdai.news/p/openai-solves-navier-stokes-metas&lt;/a&gt;&lt;/p&gt;</description>
    <category>Podcast</category>
    <category>開源模型</category>
    <category>AI Agent 基礎</category>
    <category>推論與部署</category>
  </item>
  <item>
    <title>EP153 - GPT-6 Astra 強到離譜（大部分情況）</title>
    <link>https://aiwulin.itsmygo.uk/c/5523046335/</link>
    <guid isPermaLink="false">aiwulin-5523046335</guid>
    <pubDate>Mon, 07 Sep 2026 00:00:00 +0800</pubDate>
    <dc:creator>科技浪</dc:creator>
    <description>&lt;p&gt;深入測試 OpenAI 最新發布的 GPT-6 Astra 模型，展示其在 3D 建模、機器人控制與電腦操作等領域的驚人能力。&lt;/p&gt;&lt;p&gt;原站：&lt;a href=&quot;https://www.youtube.com/watch?v=8sgNPM7cCj8&quot;&gt;https://www.youtube.com/watch?v=8sgNPM7cCj8&lt;/a&gt;&lt;/p&gt;</description>
    <category>影片</category>
    <category>瀏覽器與電腦操作</category>
    <category>機器人與具身智慧</category>
    <category>模型發布與實測</category>
  </item>
  <item>
    <title>How to Start AI Coding If You Haven’t Yet</title>
    <link>https://aiwulin.itsmygo.uk/c/2b6e735f40/</link>
    <guid isPermaLink="false">aiwulin-2b6e735f40</guid>
    <pubDate>Sun, 30 Aug 2026 00:00:00 +0800</pubDate>
    <dc:creator>The AI Daily Brief</dc:creator>
    <description>&lt;p&gt;探討非軟體工程師如何將 AI 編碼作為工作基礎技能，透過自動化、升級與發明三種建構模式，協助知識工作者解決重複、最佳化流程或創造新價值。&lt;/p&gt;&lt;p&gt;原站：&lt;a href=&quot;https://podcasters.spotify.com/pod/show/nlw/episodes/How-to-Start-AI-Coding-If-You-Havent-Yet-e3o2itk&quot;&gt;https://podcasters.spotify.com/pod/show/nlw/episodes/How-to-Start-AI-Coding-If-You-Havent-Yet-e3o2itk&lt;/a&gt;&lt;/p&gt;</description>
    <category>Podcast</category>
    <category>Coding Agent</category>
  </item>
  <item>
    <title>95% of AI Agent Projects Fail to Reach Production. Here's Why | Manoj Saxena, TrustWise</title>
    <link>https://aiwulin.itsmygo.uk/c/dab8bc2755/</link>
    <guid isPermaLink="false">aiwulin-dab8bc2755</guid>
    <pubDate>Mon, 24 Aug 2026 00:00:00 +0800</pubDate>
    <dc:creator>Eye On A.I.</dc:creator>
    <description>&lt;p&gt;Manoj Saxena 與 Craig Smith 探討企業 AI Agent 專案失敗原因，指出問題不在模型品質，而在缺乏能即時評估與管控 Agent 行為的基礎設施。&lt;/p&gt;&lt;p&gt;原站：&lt;a href=&quot;https://aneyeonai.libsyn.com/95-of-ai-agent-projects-fail-to-reach-production-heres-why-manoj-saxena-trustwise&quot;&gt;https://aneyeonai.libsyn.com/95-of-ai-agent-projects-fail-to-reach-production-heres-why-manoj-saxena-trustwise&lt;/a&gt;&lt;/p&gt;</description>
    <category>Podcast</category>
    <category>AI Agent 基礎</category>
    <category>AI 安全與治理</category>
  </item>
  <item>
    <title>Why People Are Paying 10x More for AI - and What That Means for the Chip Market | Sid Sheth, d-Matrix</title>
    <link>https://aiwulin.itsmygo.uk/c/42a5bad1c9/</link>
    <guid isPermaLink="false">aiwulin-42a5bad1c9</guid>
    <pubDate>Mon, 17 Aug 2026 00:00:00 +0800</pubDate>
    <dc:creator>Eye On A.I.</dc:creator>
    <description>&lt;p&gt;Sid Sheth 與 Craig Smith 探討 AI 晶片市場如何因「高互動性」需求而分裂，NVIDIA 架構在低延遲推理上受限，而 D-Matrix 等專用架構能提供更快的記憶體頻寬。&lt;/p&gt;&lt;p&gt;原站：&lt;a href=&quot;https://aneyeonai.libsyn.com/why-people-are-paying-10x-more-for-ai-and-what-that-means-for-the-chip-market-sid-sheth-d-matrix&quot;&gt;https://aneyeonai.libsyn.com/why-people-are-paying-10x-more-for-ai-and-what-that-means-for-the-chip-market-sid-sheth-d-matrix&lt;/a&gt;&lt;/p&gt;</description>
    <category>Podcast</category>
    <category>AI 晶片與硬體</category>
  </item>
  <item>
    <title>Why Video Agent models are next — Ethan He, xAI Grok Imagine</title>
    <link>https://aiwulin.itsmygo.uk/c/e778c76ddd/</link>
    <guid isPermaLink="false">aiwulin-e778c76ddd</guid>
    <pubDate>Mon, 01 Jun 2026 00:00:00 +0800</pubDate>
    <dc:creator>Latent Space</dc:creator>
    <description>&lt;p&gt;Ethan He 分享 xAI Grok Imagine 團隊如何在三個月內從零建立影像與影片模型，強調核心挑戰在於資料、基礎設施與人才。&lt;/p&gt;&lt;p&gt;原站：&lt;a href=&quot;https://www.latent.space/p/video-agents&quot;&gt;https://www.latent.space/p/video-agents&lt;/a&gt;&lt;/p&gt;</description>
    <category>Podcast</category>
    <category>AI Agent 基礎</category>
    <category>大型語言模型入門</category>
  </item>
  <item>
    <title>Agent-Harness.ipynb*</title>
    <link>https://aiwulin.itsmygo.uk/c/ca16a33ef2/</link>
    <guid isPermaLink="false">aiwulin-ca16a33ef2</guid>
    <pubDate>Wed, 20 May 2026 00:00:00 +0800</pubDate>
    <dc:creator>Vanishing Gradients</dc:creator>
    <description>&lt;p&gt;Vincent Warmerdam 探討 Python 筆記本如何從靜態草稿演變為與 AI 代理協作的「代理攔截器」，強調共享畫布與即時反饋的重要性。他主張使用較快模型維持開發者專注，並透過 Hooks 機制讓代理自我擴展而非依賴外部工具。&lt;/p&gt;&lt;p&gt;原站：&lt;a href=&quot;https://hugobowne.substack.com/p/agent-harnessipynb&quot;&gt;https://hugobowne.substack.com/p/agent-harnessipynb&lt;/a&gt;&lt;/p&gt;</description>
    <category>Podcast</category>
    <category>AI Agent 基礎</category>
    <category>Agent Skills 與 Harness</category>
    <category>Coding Agent</category>
  </item>
  <item>
    <title>📅 ThursdAI - May 7 - Interviews with Sunil Pai, Sally Ann Omalley from AI Engineer Europe</title>
    <link>https://aiwulin.itsmygo.uk/c/1a1e6c8fe7/</link>
    <guid isPermaLink="false">aiwulin-1a1e6c8fe7</guid>
    <pubDate>Fri, 08 May 2026 00:00:00 +0800</pubDate>
    <dc:creator>ThursdAI</dc:creator>
    <description>&lt;p&gt;訪談邀請了 Cloudflare 資深工程師 Sunil Pai 與 Red Hat 資深工程師 Sally Ann O'Malley，深入探討 AI 工程現狀。&lt;/p&gt;&lt;p&gt;原站：&lt;a href=&quot;https://sub.thursdai.news/p/thursdai-may-7-interviews-with-sunil&quot;&gt;https://sub.thursdai.news/p/thursdai-may-7-interviews-with-sunil&lt;/a&gt;&lt;/p&gt;</description>
    <category>Podcast</category>
    <category>AI 輔助軟體工程</category>
    <category>Coding Agent</category>
    <category>AI Agent 基礎</category>
  </item>
  <item>
    <title>管 Vibe Coding 项目，就像管公共厕所</title>
    <link>https://aiwulin.itsmygo.uk/c/5fbd0670df/</link>
    <guid isPermaLink="false">aiwulin-5fbd0670df</guid>
    <pubDate>Thu, 07 May 2026 00:00:00 +0800</pubDate>
    <dc:creator>AI炼金术</dc:creator>
    <description>&lt;p&gt;徐文浩與任鑫探討用 AI 管理「Vibe Coding」專案的挑戰，比喻其難度如同管理公共廁所。內容涵蓋讓 AI 治理 AI、自動修復故障、拆分任務以控制程式碼量等具體治理策略，並分享成本統計與最佳實踐。&lt;/p&gt;&lt;p&gt;原站：&lt;a href=&quot;https://www.ximalaya.com/sound/975044957&quot;&gt;https://www.ximalaya.com/sound/975044957&lt;/a&gt;&lt;/p&gt;</description>
    <category>Podcast</category>
    <category>Vibe Coding</category>
  </item>
  <item>
    <title>Andrej Karpathy: From Vibe Coding to Agentic Engineering</title>
    <link>https://aiwulin.itsmygo.uk/c/d0ad2982d4/</link>
    <guid isPermaLink="false">aiwulin-d0ad2982d4</guid>
    <pubDate>Thu, 30 Apr 2026 00:00:00 +0800</pubDate>
    <dc:creator>Training Data</dc:creator>
    <description>&lt;p&gt;Andrej Karpathy 在演講中探討從「vibe coding」到「agentic engineering」的轉變，解釋大模型如何從工具演變為具備自主行動能力的實體。&lt;/p&gt;&lt;p&gt;原站：&lt;a href=&quot;https://pscrb.fm/rss/p/traffic.megaphone.fm/CPUAI9375467036.mp3&quot;&gt;https://pscrb.fm/rss/p/traffic.megaphone.fm/CPUAI9375467036.mp3&lt;/a&gt;&lt;/p&gt;</description>
    <category>Podcast</category>
    <category>Coding Agent</category>
    <category>Vibe Coding</category>
    <category>AI Agent 基礎</category>
  </item>
  <item>
    <title>Why We Need Continual Learning</title>
    <link>https://aiwulin.itsmygo.uk/c/43b9c2b9fd/</link>
    <guid isPermaLink="false">aiwulin-43b9c2b9fd</guid>
    <pubDate>Tue, 28 Apr 2026 00:00:00 +0800</pubDate>
    <dc:creator>AI + a16z</dc:creator>
    <description>&lt;p&gt;探討當前 AI 模型訓練後即「凍結」的侷限，指出人類能從經驗中持續學習的特性。主持人與 a16z 合夥人 Malika Aubakirova 討論了情境學習（in-context learning）的邊界，並分析 Cursor…&lt;/p&gt;&lt;p&gt;原站：&lt;a href=&quot;https://ai-a16z.simplecast.com/episodes/why-we-need-continual-learning-xKmXoC4Y&quot;&gt;https://ai-a16z.simplecast.com/episodes/why-we-need-continual-learning-xKmXoC4Y&lt;/a&gt;&lt;/p&gt;</description>
    <category>Podcast</category>
    <category>研究前沿</category>
    <category>職涯與學習路線</category>
    <category>AI 編輯器與開發工具</category>
  </item>
  <item>
    <title>徐文浩：AGI 已经来了，接下来开始比赛分发</title>
    <link>https://aiwulin.itsmygo.uk/c/2e86f4a493/</link>
    <guid isPermaLink="false">aiwulin-2e86f4a493</guid>
    <pubDate>Tue, 14 Apr 2026 00:00:00 +0800</pubDate>
    <dc:creator>AI炼金术</dc:creator>
    <description>&lt;p&gt;徐文浩與任鑫探討 AGI 已實現，當前重點是產品分發而非開發。他們介紹 ZooClaw 產品，強調 Harness Engineering 以傳統工程規範制約 AI 產出，並主張從 Build 轉向 Sell，透過 Agent 協作生態解…&lt;/p&gt;&lt;p&gt;原站：&lt;a href=&quot;https://www.ximalaya.com/sound/970029642&quot;&gt;https://www.ximalaya.com/sound/970029642&lt;/a&gt;&lt;/p&gt;</description>
    <category>Podcast</category>
    <category>AGI 與長期趨勢</category>
    <category>Coding Agent</category>
    <category>Agent Skills 與 Harness</category>
  </item>
  <item>
    <title>📅 ThursdAI LIVE from London - Claude Mythos, Codex Resets, Muse Spark &amp; More | w/ Swyx and friends from OpenAI, Deepmind, LMArena and OpenClaw</title>
    <link>https://aiwulin.itsmygo.uk/c/cb05df9aa0/</link>
    <guid isPermaLink="false">aiwulin-cb05df9aa0</guid>
    <pubDate>Fri, 10 Apr 2026 00:00:00 +0800</pubDate>
    <dc:creator>ThursdAI</dc:creator>
    <description>&lt;p&gt;在倫敦現場直播，涵蓋 Claude Mythos 模型、Meta Muse Spark 及 OpenCL 安全議題。內容包含與 OpenAI、DeepMind 專家的訪談，以及對近期 AI 工具與趨勢的討論。&lt;/p&gt;&lt;p&gt;原站：&lt;a href=&quot;https://sub.thursdai.news/p/thursdai-ai-engineer-europe-mythos&quot;&gt;https://sub.thursdai.news/p/thursdai-ai-engineer-europe-mythos&lt;/a&gt;&lt;/p&gt;</description>
    <category>Podcast</category>
    <category>Coding Agent</category>
    <category>AI 工作術</category>
    <category>Claude Code</category>
  </item>
  <item>
    <title>OpenClaw 小龍蝦根本是時間黑洞！你也得了 AI 燒腦症候群了嗎？ | S2E49</title>
    <link>https://aiwulin.itsmygo.uk/c/b923a3d039/</link>
    <guid isPermaLink="false">aiwulin-b923a3d039</guid>
    <pubDate>Thu, 02 Apr 2026 00:00:00 +0800</pubDate>
    <dc:creator>矽谷輕鬆談 Just Kidding Tech</dc:creator>
    <description>&lt;p&gt;分享 OpenCL 小龍蝦作為 AI 個人助理的實際調教過程與踩過的坑，說明其設定門檻與記憶管理問題。同時探討因同時切換多個 AI 任務而產生的「AI 燒腦症候群」，並建議適時停止使用以恢復精力。&lt;/p&gt;&lt;p&gt;原站：&lt;a href=&quot;https://www.youtube.com/watch?v=KKCBbV9tVPY&quot;&gt;https://www.youtube.com/watch?v=KKCBbV9tVPY&lt;/a&gt;&lt;/p&gt;</description>
    <category>影片</category>
    <category>AI 工作術</category>
  </item>
  <item>
    <title>Why Every Developer needs to know about WebMCP Now</title>
    <link>https://aiwulin.itsmygo.uk/c/c552771160/</link>
    <guid isPermaLink="false">aiwulin-c552771160</guid>
    <pubDate>Tue, 31 Mar 2026 00:00:00 +0800</pubDate>
    <dc:creator>The AI Native Dev</dc:creator>
    <description>&lt;p&gt;探討 AI 代理如何高效與網頁互動，指出傳統截圖方式效率低且昂貴。介紹 Web MCP 作為新 API，讓代理直接執行網頁功能而非模擬點選，並提及 200MB 本地模型帶來的離線應用機會。&lt;/p&gt;&lt;p&gt;原站：&lt;a href=&quot;https://www.buzzsprout.com/2375985/episodes/18936837-why-every-developer-needs-to-know-about-webmcp-now&quot;&gt;https://www.buzzsprout.com/2375985/episodes/18936837-why-every-developer-needs-to-know-about-webmcp-now&lt;/a&gt;&lt;/p&gt;</description>
    <category>Podcast</category>
    <category>MCP 模型上下文協定</category>
    <category>AI Agent 基礎</category>
    <category>本機跑模型</category>
  </item>
  <item>
    <title>AGI is here? Jensen says yes, ARC-AGI-3 says AI scores under 1%</title>
    <link>https://aiwulin.itsmygo.uk/c/25a9ea771f/</link>
    <guid isPermaLink="false">aiwulin-25a9ea771f</guid>
    <pubDate>Fri, 27 Mar 2026 00:00:00 +0800</pubDate>
    <dc:creator>ThursdAI</dc:creator>
    <description>&lt;p&gt;Alex Volkov 與嘉賓討論 Jensen Huang 宣稱達成 AGI 的爭議，並實測 Gemini 3.1 Flash Live 的即時互動能力。&lt;/p&gt;&lt;p&gt;原站：&lt;a href=&quot;https://sub.thursdai.news/p/thursdai-mar-26-agi-here-gemini-flash&quot;&gt;https://sub.thursdai.news/p/thursdai-mar-26-agi-here-gemini-flash&lt;/a&gt;&lt;/p&gt;</description>
    <category>Podcast</category>
    <category>AGI 與長期趨勢</category>
  </item>
  <item>
    <title>GTC 2026：AI的下一个战场不是模型，而是「推理系统」| S10E04</title>
    <link>https://aiwulin.itsmygo.uk/c/4985736a6c/</link>
    <guid isPermaLink="false">aiwulin-4985736a6c</guid>
    <pubDate>Wed, 25 Mar 2026 00:00:00 +0800</pubDate>
    <dc:creator>What's Next｜科技早知道</dc:creator>
    <description>&lt;p&gt;探討 GTC 2026 上 AI 競爭焦點從模型轉向推理系統，分析 Eigen AI 等公司如何加速開源模型。內容涵蓋 GPU 與 LPU 分工、Token 消耗激增對商業模式的衝擊，以及推理最佳化技術路徑。&lt;/p&gt;&lt;p&gt;原站：&lt;a href=&quot;https://guiguzaozhidao.fireside.fm/20240417&quot;&gt;https://guiguzaozhidao.fireside.fm/20240417&lt;/a&gt;&lt;/p&gt;</description>
    <category>Podcast</category>
    <category>推論與部署</category>
    <category>AI 晶片與硬體</category>
    <category>AI 商業與創業</category>
  </item>
  <item>
    <title>Andrej Karpathy on Code Agents, AutoResearch, and the Loopy Era of AI</title>
    <link>https://aiwulin.itsmygo.uk/c/110ce419df/</link>
    <guid isPermaLink="false">aiwulin-110ce419df</guid>
    <pubDate>Fri, 20 Mar 2026 00:00:00 +0800</pubDate>
    <dc:creator>No Priors</dc:creator>
    <description>&lt;p&gt;Andrej Karpathy 與 Sarah Guo 探討 AI 代理如何改變程式開發與研究流程，分享他使用多代理協作完成任務的經驗。內容涵蓋自動實驗系統 AutoResearch 的設計理念、如何訓練模型以及未來教育模式的轉變。&lt;/p&gt;&lt;p&gt;原站：&lt;a href=&quot;https://traffic.megaphone.fm/PDP8703207384.mp3&quot;&gt;https://traffic.megaphone.fm/PDP8703207384.mp3&lt;/a&gt;&lt;/p&gt;</description>
    <category>Podcast</category>
    <category>AI Agent 基礎</category>
    <category>Coding Agent</category>
  </item>
  <item>
    <title>Is the ChatGPT Era Over? Opus 4.6 &amp; The Shift from Chat to Delegation - EP99.33</title>
    <link>https://aiwulin.itsmygo.uk/c/815fa1b3f7/</link>
    <guid isPermaLink="false">aiwulin-815fa1b3f7</guid>
    <pubDate>Fri, 06 Feb 2026 00:00:00 +0800</pubDate>
    <dc:creator>This Day in AI</dc:creator>
    <description>&lt;p&gt;訪談深入分析 Opus 4.6 與 Codex 5.3 同日發布的競爭局勢，探討百萬級語料庫、編碼模型在非編碼任務的優勢，以及代理迴圈（agentic loops）的實際成本與挑戰。&lt;/p&gt;&lt;p&gt;原站：&lt;a href=&quot;https://share.transistor.fm/s/a4bfbebb&quot;&gt;https://share.transistor.fm/s/a4bfbebb&lt;/a&gt;&lt;/p&gt;</description>
    <category>Podcast</category>
    <category>Coding Agent</category>
    <category>AI Agent 基礎</category>
  </item>
  <item>
    <title>From Swift to Mojo and high-performance AI Engineering with Chris Lattner</title>
    <link>https://aiwulin.itsmygo.uk/c/ee2b942980/</link>
    <guid isPermaLink="false">aiwulin-ee2b942980</guid>
    <pubDate>Thu, 06 Nov 2025 00:00:00 +0800</pubDate>
    <dc:creator>The Pragmatic Engineer</dc:creator>
    <description>&lt;p&gt;Chris Lattner 分享從 LLVM、Swift 到 Mojo 的語言設計歷程，解釋為何 AI 時代仍需新語言而非依賴現有工具。透過訪談了解 compiler 工程如何提升軟體效能，以及 Mojo 如何降低 AI 開發門檻。&lt;/p&gt;&lt;p&gt;原站：&lt;a href=&quot;https://newsletter.pragmaticengineer.com/p/from-swift-to-mojo-and-high-performance&quot;&gt;https://newsletter.pragmaticengineer.com/p/from-swift-to-mojo-and-high-performance&lt;/a&gt;&lt;/p&gt;</description>
    <category>Podcast</category>
    <category>AI 輔助軟體工程</category>
    <category>大型語言模型入門</category>
  </item>
</channel>
</rss>
