Now, *why* Turing-completness *wouldn't* be enough?

[Xplo]

Originally Posted by Jeffr0

Wait a minute.

Are you saying that you actually believe that humans are merely organic Turing-complete machines.

And that you can simply fold your arms... and smugly wait for someone to provide evidence to the contrary... when you provide no case or evidence for your own expansive, metaphysical claims...?

Her claims aren't metaphysical, merely physical. They're consistent with accepted science. If you're curious about her evidence, you're welcome to study human biology and psychology for yourself.

If the best counterargument you can come up with is "nuh-uh! science is dumb! you don't know things!", then yes, she can afford to fold her arms and be smug about it.

Kindly go away until you have your own explanation for how humans work and some supporting evidence for your position.

[Lamech]

If I understand turning complete machines, humans aren't one because because a) you can't run simulate another turing machine on one and b) I'm pretty sure humans include randomness via quantum mechanics.

But everything in the universe (barring some sort of super-science like souls) functions according to physical laws that can be modeled. So with a model with enough storage space, enough computing power, maybe a source of enough randomness, an accurate enough way of scanning the brain, and an accurate enough model for physics we can in fact simulate a brain.

Then when we have a simulated brain we can modify it, tamper with it, enhance it, and simplify it until we have something we feel can be called an AI. (If we aren't smart enough to do that, just find a marginally insane person and call him an AI.) BAM! And behold our new AI.

Now their might be some practical troubles along the way, like say... needing a google byte of storage space, but that is not super-science. Super-science would be having a soul that helps our brains run properly.


I also note our brains were designed by evolution, not exactly a genius inventor. The other option would therefore be to just evolve an AI. (Although that lacks the certainty of the above option, and anyone who watched terminator will probably think that is a BAD IDEA BEAR.)
P.S. If we ever get the capabilities to make an AI based of a slightly crazy person or evolve one we need to make certain its not connected to our weapons systems. Because really we would look pretty stupid after it stages a coup.

[Anthony]

Originally Posted by Jeffr0

The Chinese room thought experiement is a famous argument.

With equally famous flaws. The standard answer is that it's not the guy in the box who speaks Chinese, it's the system that speaks Chinese.

[Anthony]

Originally Posted by Lamech

If I understand turning complete machines, humans aren't one because because a) you can't run simulate another turing machine on one and b) I'm pretty sure humans include randomness via quantum mechanics.

Humans aren't a universal Turing machine. They might not be Turing machines at all, but in that case they're less powerful than Turing machines and can be simulated on a universal Turing machine.

As for QM, as far as we can tell neurons do not make use of quantum randomness, but even if they do, simply adding an RNG to a Turing machine is sufficient to deal with that effect.

[Ze'Manel Cunha]

Originally Posted by sir_pudding

I'm pretty sure that's not Ze's position. He's an atheist, AFAIK. He seems to think that the hardware is what's important and that it's not superscience to have AI; just that it's superscience to treat one set of hardware and software as a copy of another set. Or something like that. It's confusing. Like I said, I'm pretty sure he's what my Grandma calls "ornery".

Funny. ;p

My point is that algorithms can't be sentient without superscience, in a way a constructed mathematical concept, like a Code AI, is like the physical apparition of the Tardis which is a mathematical construct in Dr Who.

Building AI requires a physical construct, not a metaphysical one, or a mathematical/algorithm construct.

I'm not going to go on another go round regarding Vicky's religious views, though I do appreciate the discussions on these boards and the viewpoints raised make for good character seeds.

[Lamech]

Originally Posted by Ze'Manel Cunha

Building AI requires a physical construct, not a metaphysical one, or a mathematical/algorithm construct.

Yeah, its called a computer*. Just like any other program ever. And like any other program ever they should be transferable. Why would an AI need anything more than a computer?

*Note from this point forward I'm assuming that if needed a computer is connected to a source of randomness. Which solves any problems arising from quantum mechanics.

As for QM, as far as we can tell neurons do not make use of quantum randomness, but even if they do, simply adding an RNG to a Turing machine is sufficient to deal with that effect.

Umm... they clearly do use quantum mechanics, and which basically means that some things are in fact random. (Errors in transcription, ect.) Maybe they can all be simplified out, maybe simplifying them out will end up as a Good Thing. (No more brain cancer!) Things may seem deterministic, but their is indeed a chance that part of your brain will suddenly freeze, the random distribution of heat energy. Sure, that would be stupid to include in a model, but its their.

[vicky_molokh]

Originally Posted by Ze'Manel Cunha

Funny. ;p

My point is that algorithms can't be sentient without superscience, in a way a constructed mathematical concept, like a Code AI, is like the physical apparition of the Tardis which is a mathematical construct in Dr Who.

Building AI requires a physical construct, not a metaphysical one, or a mathematical/algorithm construct.

I'm not going to go on another go round regarding Vicky's religious views, though I do appreciate the discussions on these boards and the viewpoints raised make for good character seeds.

I never said that a mathematical construct can be recorded without a physical medium to store it on (or run it on).

[Anthony]

Originally Posted by Lamech

Umm... they clearly do use quantum mechanics, and which basically means that some things are in fact random.

I said make use of. Most neural structures are large enough that the randomness due to QM should be negligible (just like it's negligible on ICs), and randomness doesn't seem to be necessary to their operation.

[Lamech]

Originally Posted by Anthony

I said make use of. Most neural structures are large enough that the randomness due to QM should be negligible (just like it's negligible on ICs), and randomness doesn't seem to be necessary to their operation.

This disease strikes fairly randomly and I wouldn't call its effects negligible. (I'm not a doctor it might have some cause that we just can't guess at.) Although if it is the case that this is a result of randomness then yes, any simulation wouldn't include it, because that would be retarded. But the point would still stand randomness affects the brain.

[Ze'Manel Cunha]

Originally Posted by Lamech

This disease strikes fairly randomly and I wouldn't call its effects negligible. (I'm not a doctor it might have some cause that we just can't guess at.) Although if it is the case that this is a result of randomness then yes, any simulation wouldn't include it, because that would be retarded. But the point would still stand randomness affects the brain.

It doesn't strike randomly, that's a prion based disease, and as an example, the G127V prion protein mutation arose recently in Papua New Guinea tribes to insulate them from prion disease transfer vectors due to their eating of brains in their "funerary" practices.

Prions being just another part of the working structure of the brain, which would have to be modeled in the "brain-taping".