Is It Just Me?
Is It Just Me?


Is It Just Me?
You're viewing a single thread.
Early computers were massive and consumed a lot of electricity. They were heavy, prone to failure, and wildly expensive.
We learned to use transistors and integrated circuits to make them smaller and more affordable. We researched how to manufacture them, how to power them, and how to improve their abilities.
Critics at the time said they were a waste of time and money, and that we should stop sinking resources into them.
Making machines think for you, badly, is a lot different than having machines do computation with controlled inputs and outputs. LLMs are a dead-end in the hunt of AGI and they actively make us stupider and are killing the planet. There's a lot to fucking hate on.
I do think that generative ai can have its uses, but LLMs are the most cursed thing. The fact that the word guesser has emergent properties is interesting, but we definitely shouldn't be using those properties like this.
Even if you accept that LLMs are a necessary, but ultimately disappointing, step on the way to a much more useful technology like AGI there's still a very good argument to be made that we should stop investing in it now.
I'm not talking about the "AI is going to enslave humanity" theories either. We already have human overlords who have no issues doing exactly that and giving them the technology to make most of us redundant, at the precise moment when human populations are higher than they've ever been, is a recipe for disaster that could make what's happening in Gaza seem like a relaxing vacation. They will have absolutely no problem condemning billions to untold suffering and death if it means they can make a few more dollars.
We need to figure our shit out as a species before we birth that kind of technology or else we're all going to suffer immensely.
Even if we had an AGI that gave the steps to fix the world and prevent mass extinction, and generate a solution for the US to stop all wars. It wouldn't make a difference because those in charge simply wouldn't listen to it. In fact, generative AI gives you answers for these peace and slowing down climate change based off real academic work on and those in charge ignore both AI they claim to trust and the scholars who spend their whole lives finding solutions to.
Yep, those are some of the things we need to figure out before we hand our collective fate to the people in charge. I think we should start by figuring out how to keep shitty people from getting or staying in charge.