There is no such thing as "should" in evolution. Evolution doesn't do what is "best", it just follows the path of least resistance.
We are not the vehicles of evolution, just the outcomes of it. We don't owe it anything. We only exist to do what is best for us, as does every other product of evolution. In that vein, our goal in creating a better intelligence is not for the sake of creating a better intelligence, to propagate evolution. It's to create something that will be good for us and all that we value, as humans. So a superior AI that destroys humanity may be "natural" in an evolutionary sense, but it wouldn't be desirable for us, its creators. Therefore it would be a (subjectively) BAD THING. On May 4, 2014 at 9:03:14 AM, xcvsdxvsx . ([email protected]) wrote: This is going to sound radical but i dont think its immediately obvious that ai destroying man kind would even be a bad thing. Its totally natural that inferior species are overcome by superior species. Thats how life got to the point where it is today. Better designs outcompete inferior designs, the old designs die off, and the new designs take their place, only to be out-competed themselves someday. Through this process life becomes more advanced. Humans owe their very existence to this process. Without it we would still be dirt. IF ai killed us and took our place, to deny the value of this may be to deny the value of ourselves since this is exactly the process to which we owe our very own existence. On Sun, May 4, 2014 at 11:42 AM, Austin Marshall <[email protected]> wrote: I haven't read the paper (yet), but here's my take on the failsafe: For now, man poses the greatest risk to mankind. I think we're likely to destroy ourselves before a (non humanity-caused) natural disaster. If we were to build a failsafe that actually worked in AI, we'd be able to apply the same principles to ourselves. Personally, I think AI has the potential for vastly improving life on earth (and beyond). I'm optimistic that the benefits far outweigh the risk. On Sat, May 3, 2014 at 8:40 PM, Chris Jernigan <[email protected]> wrote: Well, we can add a few more names to that list of scientists who think AI will eventually kill us all. Every time someone comes out with a statement like this, I can’t help but wonder if I’m missing something. I have thought about these things over and over and over again. I just can’t see how we might get to this horrible point of no return without first implementing a fail-safe. Would it really be that unpredictable if we created a machine with true intelligence? What am I missing here? here is the original article written by Hawking, Stuart Russell, Max Tegmark, Frank Wilczek http://www.independent.co.uk/news/science/stephen-hawking-transcendence-looks-at-the-implications-of-artificial-intelligence--but-are-we-taking-ai-seriously-enough-9313474.html _______________________________________________ nupic mailing list [email protected] http://lists.numenta.org/mailman/listinfo/nupic_lists.numenta.org _______________________________________________ nupic mailing list [email protected] http://lists.numenta.org/mailman/listinfo/nupic_lists.numenta.org _______________________________________________ nupic mailing list [email protected] http://lists.numenta.org/mailman/listinfo/nupic_lists.numenta.org
_______________________________________________ nupic mailing list [email protected] http://lists.numenta.org/mailman/listinfo/nupic_lists.numenta.org
