The Jewy Nerd is not wrong that AI is getting dangerous, but he's completely fucked his mind up on the wrong detail.
I got to talking with some machine learning experts a few days ago, both of whom are pretty much under NDA so to be vague:
They told me "AI is getting really impressive in all the right ways, but to say we're a long way off from sentient general AI is like saying the Romans were a long way off from space flight"
The rationalists are freaking themselves out because AI is doing things it wasn't programmed to do and assuming that the end is near because of it. It's like Noah building the boat because it rained after a long drought. They're mistaking very interesting developments out of old algorithms for the first signs of general AI.
One guy pretty much said "If only you knew how bad things are" and that behind the scenes, OpenAI (the guys behind ChatGPT and Sydney) have some really cool tech but at the same time are running into problems scaling up past a certain point and are freaking out because they were absolutely desperate to believe this was the path to smart AI and if the results stop coming, they're going to lose investor money fast. Worse still, DeepMind (the Google team) has some internal memo that says "OpenAI fucked up one too many times, so stop publishing results online" which means that OpenAI has to stop its niggerdry of stealing everything Google makes and running off with it.
Schlomo doomered himself into a corner. He got his cult to believe that Skynet is imminent and is saying it's happening now because of language models being able to successfully fool people into thinking they're smart (you know, because they're language models; ELIZA in the 1960s did the same thing) The AIs they're working on are still fucking stupid and nothing we have even remotely resembles general AI or shows any real path to get there in the next 20 years. Worse still, this is all coming right as Moore's Law is sputtering out, and trust me, the Jews media is lying exceptionally about how useful quantum computing or graphene computers are because "computers are going to stop getting faster" is not what shareholders want to hear. The people in the know already know that general AI isn't anywhere near us. It's only the people who have nothing better to do than to masturbate to their own brain fumes that imagine that Skynet's about to be turned on because of really bad and disparate understanding of the technology.
But that's not the story you're getting from the media and the nerd hype beasts because "Skynet is imminent (and maybe we can control it)" gets more clicks and investment capital.
If Schlomo is really this autistically paranoid about general AI killing everyone, then he should be happy because we have something like another 20 or 30 years at least before we're even close to truly general AI. ChatGPT, Sydney, whatever's coming next, it's all a false start.
What he's fucking up realizing is that you don't need general AI to do a lot of crazy shit that makes it seem like general AI is near. I know DeepMind has this not-general AI thing that's probably going to get all the Singularity spergs to freak out. All that AI art shit? Holy fucking shit, you can't even begin to realize how good that's going to get really fucking fast.
It's really fucking funny how the actual engineers, researchers, experts, and data scientists keep saying "true AI is decades away" only for Singularity nerds to constantly say "Those smart people don't know what they're talking about! This rich Jew wrote an article saying that this autoregressive model shows signs of sentience and this chart says that computers are smarter than a mushroom now, which means it's coming in the next five years and is going to kill us all!" They have to scrounge together a tiny handful of people on their side (who often don't have any experience making AI) to support their claims.
It's the equivalent of an architect saying "I can't build a mile-high building on this marsh" only for a sand nigger prince to say "Yes you can, because I've seen pictures of a mile high building on this marsh and I understand some of the basics about buildings."
The real danger of AI? Dead Internet Theory and killer drones. This killer general AI shit is literal Jewish schizobait.
fredwynne.medium.com
sinceriously.fyi
www.reddit.com
deluks917.wordpress.com
