losing my mind again over how people who write scifi as a living and call it ai research have managed to convince every journalist on the planet that it's inevitable and only a matter of time before the cool text generator gains the power of god and takes over the world
like i guess in journalists' defence there's also a lot of people who really should know better who keep giving interviews like "i saw that the computer can generate text so much better than 5 years ago... therefore ive concluded that it is inevitably going to kill everyone within the next years".
there's no logical process going from one to the other, the middle step is basically "computer gains magic powers", but i have to assume that the journalists writing up these reviews go "well... they wouldn't just make a statement like that based on literally nothing, right? they're the experts after all? they wouldn't just tell me that the moon is made of cheese if it actually isn't"
also there's businesses like openai who find that, for some reason, it's really good for business to say "yes, our product will literally become god and kill everyone, and that's why you should fund our research" so they keep making those statements to get more funding
Explaining the sociology and theology of "rationalism" as Bay Area religious movement makes for a long nerdy story but it's absolutely key to understanding this. "Computers will abruptly become gods... or devils" is a whole thing and has been for about ten years and yes, Harry Potter and race science and Pascal's Wager and a staggering array of "brain enhancing" drugs are all parts of the story and I haven't even gotten to explaining what "acausal trade" is because you're going to think I'm making this stuff up
