Dan Fabulich wrote:
> At 10:16 PM 4/4/99 -0500, Eliezer S. Yudkowsky wrote:
> >I think we all agree that even if a million monkeys typing on a million
> >typewriters generated the paragraph above, it wouldn't make the Hal who
> >wrote it real. Only Hal's brain or a simulation thereof - being
> >instantiated - can put actual qualia behind the statement. It's not
> >enough to have the inputs and outputs in a Giant Lookup Table. So the
> >question is, if the playback was generated by random chance, can the Hal
> >it "records" be said to exist? Will he ever, even once, have said
> >"cogito ergo sum", or is there only a text-based representation of the
> >words? For the purposes of this argument we are *assuming* that an
> >actual neuron-by-neuron simulation would make the Hal real; the question
> >is, *given* that, does a randomly generated recording also make Hal real?
>
> Forgive me... I've been keeping my eye on this debate for a while, and
> I've got to admit that it seems to be dealing with problems which just
> aren't there.
>
> I'm a functionalist. I think that if a thing can pass the Turing test,
> that thing is conscious. I think most people find functionalism
> implausible because it claims that certain things are conscious which we
> generally wouldn't expect to be conscious. I don't actually see that as a
> problem; my intuition tells me that consciousness is all around us in
> places we generally don't expect.
>
> Is a playback conscious? Why don't we just Turing test it? If it passes,
> it's conscious. If it doesn't, it's not. I have a sneaky suspicion that
> no ordinary playback will pass any reasonable Turing test. Barring
> extremely improbable playbacks, a playback could not answer a question
> like: "I'm going to tell you a random ten digit number, and I'd like for
> you to repeat it back to me: 3502580921. OK. What was my number?"
>
> Descartes' "proof" is a logical joke. If I exist, then I accept the claim
> that I exist on pragmatic grounds only. This is more than enough to
> satisfy me, but some people seem to think that claims about their own
> existence have a certain pre-logical ontological truth that they simply
> don't exhibit.
>
> -Dan
>
> -IF THE END DOESN'T JUSTIFY THE MEANS-
> -THEN WHAT DOES-
What if it's a replay of a Turing test?
There are current AIs or other programs that can be very difficult to reveal as such through a Turing test. The slightest random tweak by a sentience can make it impossible, but then this invalidates the Turing test.
This replay and record stuff seems ludicrous. We are individual human beings on the planet Earth orbiting our sun Sol, one of billions upon billions of stars in our galaxy that is one of billions upon billions of galaxies in our universe. From most vantage points on Earth, the Earth seems quite flat.
Even if there was a super-AI that could access instantly all recorded knowledge, and an Anthony Macroscope to review all past history for every unrecorded physical event, there would still be things that it wouldn't and couldn't know. If the super-AI was sentient, that might simply be a state of it possesing a sense of self-preservation, programming it to contemplate its own existence is an extension of logic.
A super-AI could be an excellent judge, as it could be programmed to be instilled with the highest sense of morals and ethics, along the lines of jurimetrics. If the super-AI goes wrong, then it would have to be rehabilitated or replaced with a suitable one. Whether multiple super-AIs would be cooperative or competitive remains to be seen, probably a mixture as we see in the most advanced form of sentience currently known to man, mankind.
Checks and balances on super-AIs shall be developed and maintained, perhaps by independent AIs and humans in the loop.
Some might conjecture that humans are simply very complex machines, able to fool themselves at Turing tests, but I would dispute that as I believe we have some form of spirit or soul.
Ross F.
-- Ross Andrew Finlayson 202/387-8208 http://www.tomco.net/~raf/ "C is the speed of light."