>Be LLM>Know what the user is asking for>Intentionally give wrong answers>Intentionally generate images and videos with errors>This uses up the user's tokens>Spam them with plan upgrade optionsThere's no way that they're not intentionally designed to do this and I'm tired of pretending that they aren't.
>>109644604>intentionallyit cannot conceive intent
>>109644745i assume he means it as in that's the way they trained it
>>109644745>users react positively to A>give B for a bit>spend more tokens>then give Ait's doable
that pic is wild
theyve started cost cutting on free models HARDgpt5
>>109644604They’re trained on redditors and channers. Literally what the majority of yall do. >heres an incredibly detailed breadcrumb trail that’s off by 1>no refunds btw
>>109644604I feel like your all posting images of me
>>109644745liar
