A phone book for the data-science web

What it does

A single markdown file catalogs several hundred data-science and machine-learning blogs—individual practitioners, academic labs, and corporate engineering teams—with URLs and RSS feeds. It is a manually curated directory, not a scraper or ranking engine; the value is entirely in the selection.

The interesting bit

With more than six thousand stars, the repository has become a de facto community bookmark file. The roster spans from widely read names like Andrej Karpathy and Julia Evans to niche research outposts, making it a durable map of who is producing readable ML content.

Key highlights

• Mix of personal blogs (e.g., colah.github.io, karpathy.github.io), academic labs (Berkeley AI Research), and company engineering blogs (Airbnb, AWS AI, Facebook Research).
• Every entry includes an RSS or Atom feed link, implying the intended use is feed-reader ingestion rather than casual browsing.
• Plain markdown and alphabetical ordering keep it readable without any tooling.
• Community-maintained via GitHub; stale links are presumably PR-ed out over time.

Caveats

• The repository is literally a markdown list—there is no code, search, automated updating, or ranking visible in the README.
• Some entries appear truncated or incomplete in the source (e.g., “Kavita Ganesa” with no URL), suggesting the list may carry formatting debt.

Verdict

Worth a star if you are assembling a reading list or feed-reader OPML and want a crowdsourced starting point. Skip it if you are looking for a tool, framework, or algorithm.