A.I.
UPDATE - Anthropic says U.S. military can use its AI systems for missile defense ... Skynet here we come :(
Ezra Klein talks to Jack Clark about A.I.
I'm about halfway through the podcast and I'm feeling pretty depressed about the way it's going. The people creating these artificial intelligences realize that they don't completely understand what's happening, and yet they can't and won't stop or even slow down.
They tell themselves that if it can be done, it will be done. They say that if they themselves do it, at least they will have some control. But that control is ephemeral at best.
I don't know science, but I've read enough science fiction to see a possible future. I don't think I'll like it.
Ezra Klein talks to Jack Clark about A.I.
I'm about halfway through the podcast and I'm feeling pretty depressed about the way it's going. The people creating these artificial intelligences realize that they don't completely understand what's happening, and yet they can't and won't stop or even slow down.
They tell themselves that if it can be done, it will be done. They say that if they themselves do it, at least they will have some control. But that control is ephemeral at best.
I don't know science, but I've read enough science fiction to see a possible future. I don't think I'll like it.
posted by crystal | 8:50 PM
4 Comments:
I think the people developing AI have a bad case of over-confidence. I see the main purpose of it as making money. Mainly for people who already have too much of it. They talk about all the wonderful things AI will do for science and medicine. And it may actually do some of it. But follow the money. And AI is hitting resistance among people who don't want their jobs destroyed, or their electric rates to go up, or water use further diverted.
Yeah, there are lots of downsides. But it feels like it's already been decided, like it's inevitable.
This article might give you some encouragement. Turns out "consciousness" is a much more complex thing than the AI bros give it credit for:
https://aetos.ai/posts/d7e17202994c0176
The article is originally from the Atlantic Magazine, but I think it is paywalled there. Hopefully this link isn't.
Thanks for the link. I was able to read it. Very interesting. I have a fear that what might come eventually from AI will be something we can't really imagine yet. Maybe it won't need to be like a better version of us, maybe it will be something totally other that is just as real and complex and powerful as we think our nature is. Maybe even more so. And because we won't expect it, or maybe even understand it, we won't be able to survive it.
Post a Comment
<< Home