@aaronsw:
What the? Where on earth are you getting the idea that building an FAI isn't hard work? Or that it doesn't require building stuff and solving gritty engineering problems?
I'd like to reinforce this point. If it isn't hard work, please point us all at whatever solution any random mathematician and/or programmer could come up with on how to concretely implement Löb's Theorem within an AI to self-prove that a modification will not cause systematic breakdown or change the AI's behavior in an unexpected (most likely fatal to the human race, if you randomize through all conceptspace for possible eventualities, which is very much the best guess we have at the current state of research) manner. I've yet to see any example of such an application to a level anywhere near this complex in any field of physics, computing or philosophy.
Or maybe you could, instead, prove that there exists Method X that is optimal for the future of the human race which guarantees that for all possible subsets of "future humans", there exists no possible subsets which contain any human matching the condition "sufficiently irrational yet competent to build the most dangerous form of AI possible".
I mean, I for one find all this stuff about provability theory way too complicated. Please show us the easy-work stay-in-bed version, if you're so sure that that's all there is to it. You must have a lot of evidence to be this confident. All I've seen so far is "I'm being skeptic, also I might have evidence that I'm not telling you, so X is wrong and Y must be true!"
My friend, hearing me recount tales of LessWrong, recently asked me if I thought it was simply a coincidence that so many LessWrong rationality nerds cared so much about creating Friendly AI. "If Eliezer had simply been obsessed by saving the world from asteroids, would they all be focused on that?"
Obviously one possibility (the inside view) is simply that rationality compels you to focus on FAI. But if we take the outside view for a second, it does seem like FAI has a special attraction for armchair rationalists: it's the rare heroic act that can be accomplished without ever confronting reality.
After all, if you want to save the planet from an asteroid, you have to do a lot of work! You have to build stuff and test it and just generally solve a lot of gritty engineering problems. But if you want to save the planet from AI, you can conveniently do the whole thing without getting out of bed.
Indeed, as the Tool AI debate as shown, SIAI types have withdrawn from reality even further. There are a lot of AI researchers who spend a lot of time building models, analyzing data, and generally solving a lot of gritty engineering problems all day. But the SIAI view conveniently says this is all very dangerous and that one shouldn't even begin to try implementing anything like an AI until one has perfectly solved all of the theoretical problems first.
Obviously this isn't any sort of proof that working on FAI is irrational, but it does seem awfully suspicious that people who really like to spend their time thinking about ideas have managed to persuade themselves that they can save the entire species from certain doom just by thinking about ideas.