Comment by fouc

That was basically my first ever question to chatgpt. Unfortunately given that current models are guessing at the next most probable word, they're always going to eschew to the most standard responses.

It would be neat to find an inversion of that.