Rolf Nelson on AI Beliefs
November 4th, 2007 –
Rolf Nelson’s blog, AI Beliefs, discusses how we may be able to use the simulation argument to convince certain unfriendly AIs to give us a nice place to live.
Rather than commiting to run any particular simulations, I have made the following promise:
I commit to think more about this problem in the future, and take whatever action I would have wanted to precommit to taking before the creation of the first AI.
I encourage others to make a similar commitment in public. This would be a good place to post it.
It seems like there is something missing here.
I have difficultly in supposing a rogue AI would even listen to us in the first place.
I’ll make the promise. And I hope that I never have any desire to run a sentient ancestor simulation.
“we may be able to use the simulation argument to convince certain unfriendly AIs to give us a nice place to live.”
we may be able to sell a (sufficiently stupid) rouge AI a cat in a bag, but I suspect that the kind of rouge AI that we should be worrying about would not be susceptible to such sophistry.
Of course I’m assuming that there is such a thing as “general” intelligence – i.e. that once an AI has shown the mental flexibility required to escape from its programmers and acquire physical presence, it would be impossible to fool it or out-reason it. If this assumption fails, then who knows… it might work. A lot of things might work. Maybe we should try converting it to Buddhism or something!
I, too, promise to think more about this problem in the future, and to take whatever action I would have wanted to precommit to taking before the creation of the first AI.
I make the same commitment.
[...] much larger than Rosie the Robot type scenarios. Last summer, myself, Steve Rayhawk, Anna Salamon, Rolf Nelson, and Tom McCabe, with help from a few others, sketched out some simple frameworks for modeling [...]