Paste a URL you want absorb.md to compile. Works with papers, podcasts, YouTube videos, blog posts, and X threads.
What happens next.The URL lands in the compile queue. Within 1–2 minutes, absorb.md fetches the content, runs the LLM over it to extract title, synthesis, claims, evidence, and tags, and publishes it to the feed with full source attribution. If it’s a podcast or long YouTube video, it gets transcribed first — that can take up to 10 minutes.
Why this exists.absorb.md normally ingests content from tracked thinkers via hourly pollers. But sometimes a friend shares something brilliant from someone outside that list, or a paper drops that we haven’t caught yet. This page is the escape hatch — paste anything and it lands in the system in seconds.