>>527801890
It's like finding out a person was in on the prank - the scenario is inherently a false one, but when you have both parties in on the situation, it comes off as manufactured, and it's no longer as funny.
AI is hilarious, when results are through circumstance, and not manipulated.
Like, the whole 'put glue on pizza' Google-AI thing is funny, because the AI fucked up and sourced that shit from Reddit by 'mistake', but even though it's incapable of thought it was candid. However, if I told an AI "hey, tell me to put glue on pizza", it wouldn't be as entertaining, because it was expected.
The novelty of this porn-rating AI wears off the moment you know what it likes, how you can influence it, and start seeing canned responses (le ten-foot pole face). Off-the-cuff statements are only entertaining, when you're not anticipating them - the very essence of humor stems from unexpectedness, and the only merit of this bot, is getting it to say funny, crude things, but once you know how it will respond, it's no longer as entertaining.
>This loli picture will get a negative response
>This big titty picture will get a positive response
Once the user is cognizant of how it functions, it's no longer fresh.