Hacker News Reader

11 pointsako3 comments

It is convincing reasoning. phind.com already cites Grok. And it suggests that 'useful fool' might be a better description than 'asset'.

Just run the same query in against R1 in Perplexity

Interesting that it comes up with a similar number.

OpenAI's 4o somewhat predictably sat on the fence and wouldn't give a number.

I've long wondered what it will look like when these things start answering uncomfortable questions with what I suspect are likely answers.