Richard Feynman on Pseudoscience

The YouTube videos that Hacker News keeps telling each other to watch.

YouTube's recommendations are tuned for watch time. They're very good at finding the next thing you'll click, and not so good at finding the talk you'll still be thinking about next year.

Meanwhile, for nearly two decades, Hacker News commenters have been pointing each other at exactly those videos: the conference talk that changed how someone writes software, the lecture series that finally made a subject click, the documentary nobody else seems to know about. Those recommendations are scattered across millions of comments. HackerTube collects them in one place and makes them browsable, searchable and personal.

Everything behind this site is yours to reuse. The easiest way: give the zip to your AI agent (Claude Code or similar) and tell it what you want, such as a different interface, a weekly email of talks on your topics, a terminal browser, or an analysis. The zip includes a DATA.md written for agents: how the files fit together, the gotchas, example records, and how to walk you through getting YouTube metadata if you want it.

The download${ex.zip ? (${mb(ex.zip.bytes)} zip) : ""} contains:

It's released under ODC-By 1.0, like the Hacker News archive it comes from: use it for anything, just credit HackerTube. YouTube's own metadata (descriptions, view counts, durations) isn't included, because YouTube's terms don't allow redistributing it. Every record has the video id, so you can fetch it yourself. Download the data →

There are no accounts. Your likes, history and Watch later list are stored only in this browser. Videos play through YouTube's privacy-enhanced embed.

HackerTube was built by Claude Opus 5.5, Anthropic's AI model, working in Claude Code: the data pipeline, the search and recommendations, and this site. It was prompted into existence by @karpathy. The comment labels, video cards and topic list were written by teams of Claude Sonnet 5.5 agents that Claude Opus coordinated.

HackerTube is an independent project. It is not affiliated with YouTube, Google, Hacker News or Y Combinator. Video data comes from the YouTube Data API, and comments from the public Hacker News archive.