4 ms·
Seems that both authors don't really see strong AI as a possibility (at least within say next 50 years). I wonder what HN readers take on that is, but to me it
by comboy 12y ago
Seems that both authors don't really see strong AI as a possibility (at least within say next 50 years). I wonder what HN readers take on that is, but to me it seems quite likely. Not necessary designing one from the scratch, but "uploading" human brain (once we have good enough resolution and technology). It changes quite a bit about everything.
- J_ 12y agoLike you said, strong AI would change everything. The richest will be the first to utilize them. Things get scary from that point on. Unless we massively redistribute wealth, then inequality will arrive at unimaginable levels. It's hard to think of a world where the top 1% owns many multiples of what the bottom 99% owns. I'm guessing at some point that we'll decide as a species that strong AIs cannot be owned by individuals. I'm guessing that it won't be a smooth transition. Instead, we'll become quasi-communists and fairly evenly distribute gains made by the strong AIs. Also by "uploading", I'm guessing you mean simulate the human brain? Actually uploading a human consciousness would come a while after the advent of strong AI.
- comboy 12y agoYes I mean simulating human brain (of course with all advantages of digital form, including increased speed and multiple copies). I don't think that wealth would matter at all after strong AI.
- itisbiz 12y ago"Like you said, strong AI would change everything. The richest will be the first to utilize them. Things get scary from that point on." Warrants repeating / highlighting.
- wpietri 12y agoStrong AI has been 30 years away for 60 years. That's not to say we'll never get there. But it does mean that our intuitions that we're close aren't very reliable.
- J_ 12y agoNot all predictions are created equal. Those past predictions tell us that we weren't very good at predicting the coming of strong AI in the past, but it doesn't necessarily mean that current predictions are unreliable.