Recent searches
Search options
AI AITA, only slightly less credible than human AITA... https://ianbicking.org/blog/2024/05/ai-aita
Fake opinions on fake questions like...
"AITA for Asking My Husband to Stop Setting Off Fireworks Inside the House?"
"AITA for Requiring My Friends to Solve a Riddle Before Entering My House?"
@ianbicking
I always think forums like that are an exercise in creative writing rather than describing real situations. At least the stories that bubble to the top are tuned to be just believable enough to let people swallow the unbelievable bits. And the writing quality appears a lot better than what one would expect from semi random people. So using an LLM fits in quite well!
@ianbicking
I like how the LLM literally put in an asshole as the issue in one of the stories, the cat asshole. LLMs will do that.
It's impressive how the LLM has figured out the genre.
@ianbicking
I also like how DivaDevine's advice on how to completely monitor and control your husband made it through ChatGPT's alignment.
@faassen For the anti-cat reply I specifically said it should the comment should be more pro-genocide towards cats. I thought it wouldn't like that, but it went along with it fine...
Generally I find the moderation to be quite... moderate. It pays a lot of attention to context and purpose, and for something satirical like this it didn't balk at being problematic.