Karpathy 的 HackerTube 使用了 AI 数据挖掘 + 推荐系统重构 + Agentic Coding
This is a dev post classified by AGK Intelligence as AI dev tools (a tool drop), kept by the Dev Radar because it carries real work, not commentary.
Karpathy 的 HackerTube 使用了 AI 数据挖掘 + 推荐系统重构 + Agentic Coding 绕开原来的推荐机制，用另一套数据重新构建了一个 YouTube 风格的产品。先处理约 4300 万条 Hacker News 历史评论数据，再程序化粗筛：URL 提取 · Regex 正则匹配 · 去重 · 去垃圾信息，再用Claude Sonnet 语义判断，识别真正的推荐语境，过滤无效提及，最后构建推荐数据集，聚合视频 ID · 社区推荐证据 · 分类，获取视频标题、时长及其他元数据的API 里面比较有趣的是如何利用社区的长期评价信号，更准确地说，HackerTube 使用社区评论构建一个经过筛选的视频候选库，再通过网站提供发现和筛选功能。Karpathy应该使用了一套计算公式来做推荐排序。
Posted by Anita AGI/acc (12.2k followers) 2 h ago · 4 likes · 384 views · view the original post on X. Kept by the Dev Radar as AI dev tools.
More dev work like this
- 🔥还行！美团LongCat 2.5实测！3D场景端到端生成比2.5 Preview强太多 ! — @NFT_Chen
- Meet Salience-1-9B-GGUF: a multimodal vision-language model that reasons, thinks, and… — @HuggingModels
- Kinda crazy that he actually gave us both — @theo
- Turn textbook PDFs into interactive lessons—not another stack of slides. — @DanKornas
- 写专利这件事，已经有人做成了 AI Skill。 — @bkdgiffug
- Your AI can run locally and still use online tools to make its responses even better. — @TeksEdge
- OpenClaude 0.34.0 just shipped. 🌲 — @twigpine
- Day 32/ — @theo
Every post is read and classified by AGK Intelligence: what it is, which market it belongs to, and whether the link is a real tool. 31.3k posts from 5k X accounts over the last 21 days, 9k tools, 12 markets. New posts are collected several times an hour and the full ranking is recomputed every hour — last update 2026-11-02 08:27 UTC. Full method.