http://ieet.org/index.php/IEET/more/anissimov20090109/
I think his complaint is missing something. Mainly that his explanation for how things could go wrong describes something that already happens: We already have intelligent sentient beings with no morals doing their best to make money without concern for how much it costs others. Having an AI take that role won't really make things any worse than they already are.
I expect that the future for most AIs will be fairly safe, because any AI's that prove dangerous will be killed off, and evolution will handle the rest, just like how we domesticated dogs and other animals (yes I realize dogs still kill people now and then, but so do swimming pools).
The real danger is what will be done with by people who wish to do harm to others. Genocide committed by robots. Terrorism committed by robots. Etc... Humanity is it's own worst enemy, and I doubt the invention of better AI's will really change that.
No comments:
Post a Comment