About HackerTube

An independent implementation using the public HackerTube dataset.

This independent site is not affiliated with HackerTube.org, Hacker News, YouTube, Google, Y Combinator or Andrej Karpathy.

HackerTube helps you find talks, lectures, tutorials and documentaries that readers have linked and discussed in Hacker News threads. The public dataset records the context around those links, so you can see why a video was recommended instead of relying only on view counts.

The original public export spans Hacker News comments from 2007 through 2026. Videos qualify through repeated mentions across different readers and months. The dataset includes summaries, topic labels and recommendation signals; some of those labels were generated with AI and can be imperfect.

This independent implementation was written separately. It uses the public dataset with attribution, links quote excerpts back to Hacker News, and stores likes, history and Watch later items in your browser rather than requiring an account.

How recommendations are scored

The source dataset ranks videos by recurring Hacker News recommendations across separate months. Repeated recommendations count for more than a one-time spike. The score is a signal of sustained discussion, not a quality rating or a measure of YouTube popularity.

The original dataset was exported on October 31, 2026. Topic labels and summaries were generated with AI and can contain mistakes. Each available quote links back to its Hacker News discussion so readers can review the original context.

Read about the original HackerTube project · View its public data page