The game itself is, you know, fine. Anything powered by a large language model is inevitably gonna have some holes in it, and boy is this thing is making a lotta bad calls.

But I am consistently delighted by the little number that tells me how many other people guessed the stupid thing I guessed. 63 other people guessed that soap could be defeated by the “soap-eater”! And then someone else has already defeated that with poison soap! And upon defeating poison soap with antidote, yet another has already tried to defeat antidote with grapefruit juice! Which was, unfortunately, rejected by one of the aforementioned “bad calls”, but what can you do.