Mind Emerging Out of Matter via "Complexity"

[Anselm33]

Hey all,

I have a friend who argues that reason (and consciousness & free will, etc.) will eventually emerge out of computers/artificial intelligence, etc. And that, similarly, our reason “emerged” out of matter as well.

Not only do I find this vision horrifying - I also don’t agree with it. But I cannot quite say why.

Basically, what he seems to be saying is that if only substances/elements/circuits are arranged in a sufficiently complex fashion, a wholly different and independent thing emerges called consciousness/mind. No explanation is given for how this happens - much less why.

Can anyone suggest a way of articulating how and why this rather demonic vision is in error?

Thanks!

~cawbs

There are several good books refuting the notion of “strong AI”, that our consciousness is just programmable (a la Star Trek and “Data”):
see “The Emperor’s New Mind” and “Shadows of the Mind” by Roger Penrose; “Minds, Brains and Science” and “The Mystery of Consciousness” by John R. Searle. Penrose uses Goedel/s/ Turing Theorem to show that algorithms can’t produce all answers to mathematical questions (i.e. questions that mathematicians can solve) and Searle uses his “Chinese Room” argument to show that a computer isn’t aware of what it’s doing. Here’s a good article on “Mysterian” view, the “hard problem” of consciousness:
scholarpedia.org/article/Hard_problem_of_consciousness
consc.net/papers/moving.html
And from J.P. Moreland’s book, “Consciousness and the Existence of God”, a blog: afterall.net/papers/490852
Hope these are helpful.
Anselm.

 

[kildare]

Sorry Touchstone I understand what you and innocente asked, but I’m not convinced, Not by a long shot and not even if you had the greatest supercomputer that could possibly exist.
Have you ever tried modeling non-linearties? What about something relatively simple like the Navier-Stokes equations for Turbulent flow. You able to predict the precise location and velocity? My problem is that you need all the information and more. You need your electrons to act like mathematical points in space, and they don’t even do that. Heisenberg is actually relevant to this I’m afraid.

You know what your unfounded optimism makes me feel very contemptuous and that makes me sad, because its not a christian feeling at all.

So anyway, lets entertain your preposterous idea for a minute. Our IBM super computer can undoubtedly clone copies as well. Maybe your willing to take a risk on the tele-transporter. Lets just say that I feel a bit skeptical about being de-materialsed into data and then being re-made. I just don’t buy it, that that will be me. But then Scottie on the Enterprise presses the copy button by mistake again. He copies you twice. Which body, or which mind, do you now inhabit if any? Are those perfect clones really you?

At the end of the day the Judeo-christian belief is not pantheistic nor is it dualistic. The Body and the soul are connected as one. Only God can make the perfect copy and we are never, in this world at least, going to know how that is done

 

[kildare]

I tried giving them that answer, though no where near as detailed as you just did, they refuse to accept it, such is the love for that goofy thought experiment, not even getting into the fact that it ends by asking if the silicon brain would be Turing machine, a theoretical device that reads marks on an infinite strip of tape

Well brother, we all like to hold on to our beliefs I suppose, and I think they are trying to be open minded, but they wont see it until they accept the Lord. The important thing about the atheist psychology, it seems to me, is that everything be made reducible so that we can understand it, like having electrons as planets so that we can conceptualize it. Anyway if something can be understood and dominated, it feeds the ego, and thats were the real problem is.
Humility is the opposite and the first step to wisdom.
The greatest scientists were actually quite humble.

 

[Spock]

Well, here we depart. If you don’t have a way to distinguish, as you allow right here, you don’t have a basis to agree, do you?

I am not sure what you mean here. I don’t have the basis to agree to what?

Let’s do another thought experiment. Suppose we have a perfect copy machine, which reads every atom and is able to put together a perfect replica of the original. Suppose we put the Mona Lisa into the machine and get a copy. Then put both picutres on the wall, side by side. Call in all the experts and ask them which is the original. Obviously, one of the pictures was touched by the hands of Leonardo and the other one was not. In that sense, there is a difference. But that difference cannot be measured. My stance is that there is no difference between the two. The distinction of “original” vs. “copy” is not applicable if the difference cannot be measured.

It’s not “clearly” unconscious, even (or especially) as a matter of brute force processing. That’s very much what may be at issue. Humans themselves are a challenge because of their “brute force” advantage over modern computing machinery; it’s so massively parallelized and asynchronous that no man-made machine at this point is practical as an analog. It would be unthinkably complex, not mention impractically huge, since our best small-scale manufacturing is no match for the fine grained nano-biology of the human brain.

Sure thing. The current techonology is not “good enough”. But we talk about a hypothetical scenario. Let’s say that we put a person into the copy machine. How do we tell afterwards which was the original and which is the copy?

Thinking is as thinking does. Emotion is as emotion does. Love is as love does. There’s nothing inherently silicon or organic about that that we can identify. Rather these are descriptions of complex patterns and phenomena, the interrelation of complex systems.

I agree.

 

[inocente]

How would we program human imagination, human morals, human pleasures, human dilemmas, etc. into a computer?

And why would we want to or need to, when we already have them in ourselves?

Creativity, imagination, is fairly easy to program - just do random lookups of concepts and throw them together until something emerges. Young people are often more creative because they haven’t learned enough common sense yet to suppress the more outrageous combinations. A crude way of programming this is an evolutionary algorithm – create “bots” randomly, pose some kind of survival task and see which survive.

But a computer probably couldn’t simulate human thought without our sensory (name removed by moderator)uts, body shape, societies, etc. Philosophically I think this is known as the What is it like to be a bat problem. The computer (which would be a rather demeaning term for such a sophisticate beast) would see things its own way.

 

[inocente]

I´ll stick my neck out and say it cant be done - that is you cannot perfectly replicate the brain, and my feeling is that is due to inherent non-linearity of each component that makes up teh whole. Each nueron is unique in that it might have variation of density, fluctuations of forces. Perhaps even at the aub-atomic level there are variations.
Then there is chaos theory. You know the famous butterfly effect. Well, it means that even allowing the tiniest of inacuracies in the initial conditions and you might get wildly different results.
Now if there are variations at the sub-atomic level, and if we need to iron those out two, then we are alread yundone, because we are already entering the Heizenberg Uncertainty Principal area.

Thanks for having a go . Don’t agree though but you saw that coming. True it can’t be done, that’s why it’s a thought experiment. The purpose is to try to identify x-factors, and you’re proposing chaos, Copenhagen QM interpretation (perhaps) and Heisenberg.

Chaos is dealt with by replicating all the gate values precisely (ronnie’s alternative biology experiment seems to have the same difficulty here). However the brain inherently has the ability to inhibit feedback and run-away amplifications (except perhaps temporarily in epileptics) so after a settling process we’re neatly back on track.

The QM interpretation (if you’re suggesting it) is already known to be wrong – any stray electron that comes close to a quantum computer “observes” it and destroys the qubit, so the brain would have the same problem in spades.

Using Heisenberg would assume that large-scale neurons consisting of loads of atoms can be affected in some unspecified way by sub-atomic phenomena and doesn’t really seem like a starter.

However, it’s encouraging that your x-factors are all physical.

Sorry Touchstone I understand what you and innocente asked, but I’m not convinced, Not by a long shot and not even if you had the greatest supercomputer that could possibly exist.

Well brother, we all like to hold on to our beliefs I suppose, and I think they are trying to be open minded, but they wont see it until they accept the Lord.

Now hang on bro, am I putting 2 + 2 together and making five or are some people (there is another) around here claiming I’m not a Christian? In which case just because some Catholics may hold to a particular worldview doesn’t advance a right to say they know what others believe .

 

[inocente]

As an “emergentist” (aka non-immaterialist, materialist) I’m trying to understand how an immaterial architecture of mind works. So only half-jokingly:

Definitions - the mind consists of a material part (the mUnit) and immaterial part (the iUnit) necessarily connected via an interface (the miFace ). The mUnit is the entire nervous system including the brain while the architecture of the iUnit and miFace are unknown, although a few things can be deduced.

The miFace must presumably be spread over the whole of the mUnit, since if it were localized we’d have detected it by now. Also of course, a localized miFace might fail in cases of brain damage leading to allegations of non-humanness. Presumably explaining the continued functioning of a non-local miFace while the mUnit grows from baby to adult entails the idea of one or more essences.

The iUnit can’t provide a net inflow of information over the miFace since that sounds like magic (taking over another’s soul, real predictions from chicken entrails, etc.). The iUnit must therefore be a processing adjunct, manipulating data on behalf of the mUnit.

The mUnit’s emotional and intellectual processes will probably be fully explained in entirely material terms before the turn of this century, so to remain valid the immaterial explanation may limit the iUnit to only the most high-level and important subjective feeling – let’s say awareness of self. Thus even though Touchstone’s prosthetic brain reports the awareness and behaves in all respects as if possessing the awareness, its awareness can be claimed as not authentic.

As an emergentist this remains a moral question for me – in the face of increasing knowledge there will be a point fairly soon where neurosurgeons routinely make changes to the mUnit, including perhaps adding silicon, and the immaterialist must face the question of what it is to be human, while the emergentist can just repeat the mantra human is as human does.

 

[Charlemagne_II]

I can see how a computer can be programmed to laugh. What I cannot see is how you program a computer to see a reason to laugh.

 

[inocente]

I can see how a computer can be programmed to laugh. What I cannot see is how you program a computer to see a reason to laugh.

To help you, first-off recruit some middle school kids who understand the materialist explanation of laughter – they get taught it these days – sciencenetlinks.com/lessons.php?DocID=381

Then you’d have to surprise and delight the computer and the kids, for example by telling them some nonsense such as the immat … well you can fill in the rest for yourself.

 

[Charlemagne_II]

inocente

If you joked about computers to a computer, would the computer know to laugh along with you?

 

[Spock]

If you joked about computers to a computer, would the computer know to laugh along with you?

If I were to tell you a pretty sick, but hilarious joke about Christianity, and you would not laugh along, would that prove that you are not a rational being? Maybe you would not find it funny, or maybe you would not even understand it. Well, the computers - TODAY - would not understand it. DUH!

I recall another joke. Bush senior’s vice president, Dan Qualye is apporached by a journalist, who asks: “Mr. Vice President, are you offended by the jokes about you?”. Qualye answers: “No, of course not”. Then the reporter asks: “And when your kids explain them to you?”.

 

[inocente]

inocente

If you joked about computers to a computer, would the computer know to laugh along with you?

Except for speciesist humor , when it would get morally indignant and “blue screen” bless its little cotton socks.

 

[Charlemagne_II]

Spock

Well, the computers - TODAY - would not understand it. DUH!

But the computers tomorrow will understand it?

 

[inocente]

But the computers tomorrow will understand it?

If and when humor has a functional purpose for them, e.g. evading the topic.

 

[inocente]

But the computers tomorrow will understand it?

If and when humor has a functional purpose for them, e.g. evading the topic .

 

[Charlemagne_II]

inocente

*If and when humor has a functional purpose for them, e.g. evading the topic *

If and when they get a soul … never.

 

[PeteO]

Now someone may argue that this system will only “emulate” the biological brain. We must ask them, what is the difference between a “sufficiently convincing” emulation, and the “real McCoy”? If a computer will pass the Turing test, on what ground can anyone consider it “inferior” to a real human?

You posit a very interesting concept! I find this a fascinating discussion- but also quite revealing about our psychology as humans.
This might be an insight into how we should perceive ourselves in relation to God!
We would probably always view artificial intelligence as inferior due to the fact that it was just what we say it is-
Artificial.
Created things are always viewed as lesser (philosophically speaking) than the being that creates them!
The limits of artificial intelligence will reflect our limitations as human beings.
For example, no matter how much animals mimic human intellect, they do not possess the understanding capacity of human beings.

No matter how much we humans are able to program an artificial intelligence to mimic our understanding- we are not creating anything new- we are creating an intelligence in our own image, with our own logical structure, with our own limitations on thought, etc. The intelligence will only mimic ours.

We can certainly program a machine to pass a Turing test- yet it will only philosophically speaking be the PROGRAMMER that passes the tests. Not the machine.
There would be no good way to prove the machine as having self-awareness. Heck, I can program a website to claim it is self-aware.

It reminds me of a scientific american article I once read when I was little, that claimed that scientists created a robot that could fix itself! It could build OTHER robots! I thought we had attained self-aware AI.

Then when I carefully read the article- it described how the robot did such things- and it turned out, it’s actions were predictable in many ways, and unpredictable in some. But only in the ways that we programmed it to be. Alas, no self-awareness… no true free will…

 

[inocente]

If and when they get a soul … never.

If by soul you mean the spiritual, the part which experiences deep feelings and emotions, we’re back to carbon vs. silicon brains and how we could know it has these subjective feelings.

Assume for some strange reason we set out to make a silicon entity in our image, how would God evaluate it? Worthy of immortality? Similar kind of question to whether an intelligent alien species could be in God’s image.

I’d suggest the elephant in the corner is still this moral question of who of us here below has any right to decide.

 

[Spock]

We would probably always view artificial intelligence as inferior due to the fact that it was just what we say it is-
Artificial.

This is an understandable approach, as long as you consider the current stage of technology, and only concentrate on the available protheses. Quite obviously the current artifical limbs (for example) are less sophisticated than the original ones. A machine which acts as a kidney is large and cumbersome. An artifical lung is a big box, surrounding the patient. Artifical eyes are just a cruel joke, mimicking the looks of the eye, but not the functionality. So most of these protheses are hugely inferior to the original - as of today - that is true.

On the other hand, a car or a plane are seriously superior to the legs, as means of transportation.

Created things are always viewed as lesser (philosophically speaking) than the being that creates them!

I disagree with this, and see no reason to accept it. And I have no idea what you mean “philosophically speaking”. A child - created by the parents, can be smarter, stronger, better, kinder, more talented, etc… than the parents.

No matter how much we humans are able to program an artificial intelligence to mimic our understanding- we are not creating anything new- we are creating an intelligence in our own image, with our own logical structure, with our own limitations on thought, etc. The intelligence will only mimic ours.

Not necessarily. The artifical means of transportation do not mimic the movements of the muscles in our legs.

We can certainly program a machine to pass a Turing test- yet it will only philosophically speaking be the PROGRAMMER that passes the tests. Not the machine.

Not true again. There are programs even today, which evolve, which get “smarter”, which are much more complicated than the original used to be. The programmer, who did the original design cannot even begin to comprehend the complexity of the program after a time.

There would be no good way to prove the machine as having self-awareness. Heck, I can program a website to claim it is self-aware.

I am sure you can. But “claims” are dime a dozen. Can it pass a rigorous Turing test lasting for a few weeks?

 

[Touchstone]

I am not sure what you mean here. I don’t have the basis to agree to what?

ronnie bonigli said:

In the Emperor’s New Mind Penrose brings this up and points out that we could program the computer to give occasional “flawed” answers, so it’s “more human sounding”. With enough time and subroutines IBM could build a system that fools anyone yet was clearly no more conscious than your PC.

You replied:
*

You see, I would agree with that, but how do you know it?

Such a system is NOT “clearly no more conscious” (perhaps you’re “agree” wasn’t referring to that, but something else ronnie said?). It may be just as conscious in the sense that it “fools anyone” in the same manner humans do with their minds - fast, parallel processing from a huge and ever-updating database. Such a machine may NOT be conscious, but there’s basis for saying it is “clearly no more conscious”. It might be a close facsimile to the model cognition model.

Let’s do another thought experiment. Suppose we have a perfect copy machine, which reads every atom and is able to put together a perfect replica of the original. Suppose we put the Mona Lisa into the machine and get a copy. Then put both picutres on the wall, side by side. Call in all the experts and ask them which is the original. Obviously, one of the pictures was touched by the hands of Leonardo and the other one was not. In that sense, there is a difference. But that difference cannot be measured. My stance is that there is no difference between the two. The distinction of “original” vs. “copy” is not applicable if the difference cannot be measured.

Totally agree. And that’s why I was surprised you agreed with what ronnie said, as that principle was what he was rejecting. A machine that creates perfect copies (however that is functionally measured) doesn’t have a meaning functional difference. Silcon or water, the output and performance is what matters. When they match, they match.

Sure thing. The current techonology is not “good enough”. But we talk about a hypothetical scenario. Let’s say that we put a person into the copy machine. How do we tell afterwards which was the original and which is the copy?

I don’t think we can, based on the stipulations we make about this machine. It’s tautological on a straightforward reading. Without something that defeats functionalism (“God knows each soul”, etc.) the two are definitionally identical.

We’re thinking along the same lines here, generally. I think there’s often a tendency even for AI proponents and materialists to dismiss “brute force” as a/the substrate of intelligence. Ned Block’s Blockhead objection to the Turing Test is interesting but I think more problematic in its own right than the Turing Test. Any human-like intelligence will bring a enormous amount of brute force, massively paralllel processing to the table. That is likely not all it entails, but that is how the superstition and magic becomes demystified; it’s just computing machinery at massive scales, parallel-wise and in sheer number of computing nodes. Humans appear, from science, to be a lot more like Block’s Blockhead, architecturally, than Block understands.

-TS