Hacker News Reader

It’s just a text completion model. It doesn’t actually evaluate computations, despite it being common to find it giving correct answers for some computation queries

View on HN

That's true. It does not acknowledge this limitation when communicating with the user, even when I prompted the model with suggestions.

Reply on news.ycombinator.com