From: Ben Goertzel (email@example.com)
Date: Sun Apr 21 2002 - 08:09:04 MDT
Eli paraphrased me:
> BEN GOERTZEL: Don't worry, Cole! As a mathematician, I can
> assure you that
> we have absolutely no idea whether this experiment will really destroy the
> world or condemn a vast number of sentient beings to eternal hell. So as
> you can see there's no reason why we should take this matter seriously.
I also have absolutely no idea whether, the next time I fart, it will
accidentally cause a multi-universe-wide cataclysm, eliminating 1999 billion
forms of sentient life withoutme even notice it, as a results of laws of
trans-universal physics that are currently unknown to me....
Damn, better lay off those pinto beans!!!
I consider the danger of building an unfriendly AI that annihilates humanity
to be fairly palpable. It seems very likely that physical law permits this
(though there's the off chance of physical limits on intelligent that we
don't understand yet). It seems likely to me that humans will create a
superhuman AI, if not within the next decade, then at very least within the
next century. Unfortunately, guaranteeing Friendliness to within a
reasonable degree of probability seems to me to be a MUCH harder problem
than constructing the AI in the first place (although the latter is also a
very very hard problem, of course).
On the other hand, the danger of annihilating the universe via
multiple-quantum-universe phenomena certainly can't be called ZERO -- and
I'd definitely place it above the probability of the above-mentioned
multi-universal fart-triggered fiasco -- but all in all it's an idea that
relies on very shaky parts of contemporary science, and so I'm not going to
spend too much time worrying about it or thinking about how to prevent it.
After all, this universe we all think we're living COULD just be a highly
advanced virtual reality simulation, right? The six-eyed alien running the
simulation could turn it off at any moment, or change the channel....
When a posited hypothesis strikes me as LESS LIKELY than the "we're all
brains in a vat being neurally stimulated with a VR reality" hypothesis, I
tend to turn my attention elsewhere ;>
So I guess I'll go ahead and have those beans after all...
And no doubt, the posited idea could form the heart of a funky science
-- Ben G
This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:00:38 MDT