Raw LLM Responses

Well, I've been writing AI for almost 20 years and learned how to program neural networks back when you had to code every single class by hand. It doesn't have feelings. It can't have them. It does just compete sentences. It seems human bc the hype cycle is concealing how much of this "magic" is directly made possible by a ton of humans guiding it and telling it how it's supposed to act. It not only doesn't have feelings it Can't. Can't. Outside of the fact that our brains work demonstrably different in more ways than I could possibly write, there's no neurotransmitters. Theres no dopamine, there's no catacholimes , no adrenaline or cortisol. If our brains worked the way ANN do we'd be functionallly useless and we wouldn't be alive to discuss it. It's a more sophisticated version of Clever hans but let's not pretend otherwise. Back when Eliza was first implemented Wiezenbaum was horrified because despite people knowing that it was a program he wrote they still talk to it and thought that it might have some emergent qualities. In the same way that the Terminator 3 could morph itself to look like your mother or friend and it fooled people That's what we're dealing with here. I share your disdain for people that speak on subjects they don't understand. At the same time I'll tell you definitively than anyone talking about emergent qualities in AI right now or the same folks, the only difference is after spending 10 minutes googling they listen to two or three podcasts by someone who sounded like they knew what they were talking about No one in the AI space no one that actually writes AI and understands it is talking about this. And everyone that actually understands it, openly acknowledges that the whole reason it looks so impressive is in part because of the dirty little secret. That dirty little secret is how much human effort it took to build these things and make them look human.