Hacker News Reader

Curious: What are the trivial agent vs. human detection mechanisms?

I don't know, but the article indicates that sites looking to distinguish are doing so at something like 96% success rate.