This is exactly the type of thing a bot sucks at. Training image recognition or text recognition is trivial. Training concepts such as perceived temperature requires state of the art LLMs.
This isn’t the only question, of course. It’ll stick around until bots are trained for temperature. Then it’ll ask “which of these would make a good picnic spot” or “which of these are big enough to contain an apple”.
FuglyDuck@lemmy.world 1 year ago
Turing tests aren’t just about knowing things that a human would. Bots giving a perfect answer where humans would struggle is a perfectly acceptable way to filter.