TLDR

How this community works

What belongs here

High-signal discussion of AI: news, research, tools and what it is like to build with them. Post what you yourself would want to read, ideally something a person deep in AI would find interesting.

Rules

  1. Comments must be written by you, not an LLM.
  2. Submitting your own work is actively encouraged as long as it is a small share of what you submit.
  3. Treat people with respect. Disagree with the claim, not the person.

What a good comment looks like

Reflexive cynicism, snark, empty hype and self-promotion make the discussion worse.

Scoring

Every comment and every submitted link is scored by an LLM judge that gives it an initial score. The score reflects how much the post adds to the discussion.

Over time, more and more of the scoring is weighted towards upvotes and downvotes, so the LLM judge is primarily responsible for a comment or thread’s starting position.

This is different from traditional upvote and downvote forums like Reddit, and has two major advantages:

We treat submitting links as a separate privilege. We give new members submission privileges after they have posted at least one high-quality comment.