Hacker News Reader

If it's trained on material scraped from the web, I imagine it would include all the test prep sites and forums.

Could they not have easily excluded any page with terms like LSAT? I’m sure it wouldn’t catch everything but it would probably be close.