Forming your own thoughts because you reasoned by yourself?
AI just goes “I’ve seen X before, someone answered Y, therefore I will answer Y.” In its current state it can’t decide “I’ll answer something nonsensical just for the lulz” because it doesn’t know if Y is right or wrong, it just knows that over billions of lines of texts it has seen X with Y most often so X = Y. If X was always answered with a nonsensical answer it would repeat it even if it has access to information that proves that answer wrong. Which is also why there’s a lot of bad info being shared by AI.
Forming your own thoughts because you reasoned by yourself?
AI just goes “I’ve seen X before, someone answered Y, therefore I will answer Y.” In its current state it can’t decide “I’ll answer something nonsensical just for the lulz” because it doesn’t know if Y is right or wrong, it just knows that over billions of lines of texts it has seen X with Y most often so X = Y. If X was always answered with a nonsensical answer it would repeat it even if it has access to information that proves that answer wrong. Which is also why there’s a lot of bad info being shared by AI.