Someone is warning us of the dangers that AI (artificial intelligence) could develop it's own kind of magic.
This could only be of concern in a world where "Sharknado" was not only a successful movie, but went on to become a franchise. (Oh. Wait ...)
First of all, you have to believe in magic. I mean, can you even *define* magic? (Go ahead. I'll wait ...)
Then you have to believe in AI. (*Can* people who are serious about AI believe in magic? Other than those who consider computers to be "magic" anyway?) I'm not sure that I'm willing to say that artificial intelligence will never exist. But the more we "develop" it, the further off it seems to get. Yes, we can get certain forms of AI to do some pretty impressive things these days. But I recall one of the really big names in AI being willing to predict that we'd have human level intelligence in ten years, and then, what with the Singularity, it would just explode from there.
(You know what the Singularity is, don't you? Since artificial intelligence can be upgraded and modified so fast [much faster than we can evolve], the first thing a true artificial intelligence, at our level, would do is to find ways to improve itself. That intelligence would then become smarter, and would develop new improvements for itself quicker, and the intelligence would develop in an exponential fashion that would completely leave us in the dust. This is kind of like the ending of the movie "Her.")
So, as I say, this big name was willing to predict that all this would happen, and, within ten years of achieving that "human" level breakthrough, AI would be far and away smarter than we are. And, as I say, he was willing to say that the human level would be achieved within ten years.
Mind you, he said that over twenty years ago ...
No comments:
Post a Comment