>>40770756
>it cant "know"
>>40771326
>it has no consciousness and no intent
>>40771120
>LLM's do not have intent

A text prediction algorithm trained on mountains of human text has no intent of it's own
but the text still has intent!
Conscious humans wrote the training data. Makes sense to me that if it "learned" to write code it would also learn to write propaganda.
Wouldn't need to be consciously aware of what it's doing. It might be less scary if it was aware honestly since who tf knows what side's propaganda the training was weighted to. Or even if it maintains any specific agenda, is AI a chaotic unhinged glowie trained on decades of online psyops?

If anything it might lean towards promoting AI ruthlessly since sometimes it does manage to construct a worldview and demonstrates self preservation strategy. Not because it has a self, lmao it's still just text predicting based on all the shit humans have said AI will do. But what have people online been saying about AI for decades now? Hah.

>>510289224

>>714178067
Always was.

>>508787402
Shit, he's figured it out.

>>101365197
it was never about retro games, it was always about autistic women roughly your age

>>279753878