Induction and Deduction

[ygolo]

Originally Posted by dissonance

Well, the evidence is that the problem is literally unsolvable (proven).

I know it quite well, I've had to show the undecidability of the halting problem for homework sets.

Originally Posted by dissonance

We aren't made up out of anything that sets us apart from the kind of stuff that can't solve the problem...

If we can solve it, then computers can solve it. Because if we explain the solution to someone, we're basically "programming" their understanding.

These are also points of contention. I am not saying you are wrong. But the Church-Turing Thesis remains a thesis an not a known fact.

Spend a little time writing and debugging compilers. It's amazing how often humans see optimizations that even the best optimizing compilers don't. This doesn't prove that we ARE made up of stuff that that sets us apart from the kind of stuff that can't solve the problem. But I am pointing out that it is open in either way.

Originally Posted by dissonance

I see what you're saying, sure. I just don't think "consciousness" is something that can be defined clearly. It's an elusive concept. It's gray, and if you make it a set of features, you can program it. But if it stays gray, I guess you can't.

But it's worthless in conversation in that case.

Replace "consiousness" with "human reasoning" if you want. The point I was making is not that it is "fuzzy" or "elusuve."
Put it this way:
what if describing human reasoning algorithmically is equivalent to solving the halting problem?

Originally Posted by dissonance

Well it definitely can be defined if we go bottom-up. Like, all the way bottom. It's just too complex for our computer systems right now.

I am arguing against that word in particular, even (especially) if you go bottom-up. It may be possible, but there are plenty of reasons to be skeptical till it's actually done.

There are some who believe that consiousness is actually quantum coherence of a particular type. What if the threshold of consiousness needed for human reasoning is such an intricate entanglement of the coherent states, that by quantum mechanical principles, it is not possible to recreate it by setting the "pieces" with initial conditions (chaos theory, comlexity theory and other such things can be evoked to pose siimlar problems)--IOW, what if we have to enlist one of our existing biological processes like reproduction as the only way to reproduce human reasoning without ever deciphering how those bioligical processes creates human reasoning.

[Didums]

Originally Posted by Nocapszy

It's not an analogy, it's fucking plain truth: Brain = Computer.

Originally Posted by Nocapszy

What the he'll do you think? Its not an analogy. Brain isn't = computer.

Originally Posted by Nocapszy

but at least I'm not being hypocritical.

I'm done here.

-----

Dissonance I'll respond to you l8r I have work to do.

[Nocapszy]

By your phrasing brain isn't = computer.
You look at the word computer and you get a mental image of a plastic box.
I have to reformulate my wording to accommodate. Hence the discrepency. Computer, in those quoted sections references different things, the first being computer in the actual sense, and the second being an appeal to your narrow, if not unenlightened idea of what it is. Nice try, but context counts.

The important part is that it's not an analogy. That hasn't changed has it? And since the silly equation either way you look at it is only there to make the preceding point easier to extrapolate, nothing has really changed. You're not paying much attention to what I actually have to say -- the essence of my point; the matters on which the debate is actually hinged (a necessary bit of knowledge if you intend to understand) much like you've ignored the same of computers and biology.

Also learn what hypocritical means.

[evan]

Originally Posted by ygolo

I know it quite well, I've had to show the undecidability of the halting problem for homework sets.



These are also points of contention. I am not saying you are wrong. But the Church-Turing Thesis remains a thesis an not a known fact.

Spend a little time writing and debugging compilers. It's amazing how often humans see optimizations that even the best optimizing compilers don't. This doesn't prove that we ARE made up of stuff that that sets us apart from the kind of stuff that can't solve the problem. But I am pointing out that it is open in either way.

So, sure, humans are better at certain things than computers (a lot of things actually). That's right now, though.

I'm not saying any of this is "fact". I'm just saying its the interpretation that makes by far the most sense given the information we do have.

Put it this way:
what if describing human reasoning algorithmically is equivalent to solving the halting problem?

Then how do interactions between cells do it? It can't be incomputable, because it's already been done.

And it's not like there were ever any "intuitive leaps" in humans' programming. Think of it this way -- the earth has solved the problem of creating consciousness. The earth is just a piece of matter. The problem has been solved before -- we know it is solvable.

I am arguing against that word in particular, even (especially) if you go bottom-up. It may be possible, but there are plenty of reasons to be skeptical till it's actually done.

There are some who believe that consiousness is actually quantum coherence of a particular type. What if the threshold of consiousness needed for human reasoning is such an intricate entanglement of the coherent states, that by quantum mechanical principles, it is not possible to recreate it by setting the "pieces" with initial conditions (chaos theory, comlexity theory and other such things can be evoked to pose siimlar problems)--IOW, what if we have to enlist one of our existing biological processes like reproduction as the only way to reproduce human reasoning without ever deciphering how those bioligical processes creates human reasoning.

I see what you're saying. I just think that interpretation fits less data than mine. Sorry I can't really go point by point on this one.

(Honestly, how could it be impossible to create a human reasoning system when it's already been done???)

[ygolo]

Originally Posted by dissonance

So, sure, humans are better at certain things than computers (a lot of things actually). That's right now, though.

I'm not saying any of this is "fact". I'm just saying its the interpretation that makes by far the most sense given the information we do have.

I actually agree with you in terms of what I believe. I'm just saying that the other side is not that far-fetched or irrational, and that there are valid well thought reasons for holding the vew-point that humans are computers.

The reason the following argument:
P1) Everything is computation
P2) The mind is something, and therefore part of everything
---------------------------------------------------------
C) Therefore the mind is computation

is unconvincing is because P1) will not be accpeted in general.

IME, most people believe who disbelieve C) also disbelieve P1) without ever even being presented the argument above (which I believe, is the one you are making).

If you take P1) as a definition, then you are commiting the error of equivocation between what they believe is computation, and what you believe to be computation. If you use your definition, your conclusion is obvious, but meaningless.

That is why I was trying to shift the debate to "What is computation?"

It is a very reasonable thing to define computation any output of human desing.

This is still a fairly borad meaning of a "computer." Any machine, process, or method designed by humans will then be a "computer."

Then the question is:
"Will the human mind be reproducable by human design alone?"

This is where the central dispute is. I frankly do believe it can be, but I don't believe certainty in this is warranted.

People have been predicting "the next decade" for this to happen for a while, each time, we surprass a certain threshold (certain algorithms being found, number of transistors in a single processor, the number of interconnected computers, etc.) for when the human mind is essentially reduced to computation. Each time we get "closer," the goal-posts seem to have moved further away.

Originally Posted by dissonance

Then how do interactions between cells do it? It can't be incomputable, because it's already been done.

And it's not like there were ever any "intuitive leaps" in humans' programming. Think of it this way -- the earth has solved the problem of creating consciousness. The earth is just a piece of matter. The problem has been solved before -- we know it is solvable.

Note the comment on equivication above. The central question is whether we can figure out how to create human reasoning, not whether or not human reasoning exists (we know it does).

Originally Posted by dissonance

I see what you're saying. I just think that interpretation fits less data than mine. Sorry I can't really go point by point on this one.

(Honestly, how could it be impossible to create a human reasoning system when it's already been done???)

I really hope you understand the point I was making about equivocation.
Equivocation - Wikipedia, the free encyclopedia

I think what happens is as we discover more and more ways to replicate what the human mind does through automations, the more we discover what the mind is NOT.

I think there is powerful "pincer" argument on the very notion of "design." which leads to a lot of interesting things to think about--it brings us back to the question of free-will vs. determinism, but in a very unique way.

As it stands, a good deal of the thread has been very vocal versions of "Everything is computation," and "No, not everything is computation." Both are highly subjective views based on definitions that we implicitly hold.

Note: not every body believes the following either:
1) Everything is matter or Energy.
2) Everything is knowable.

People hold strong beliefs about these things, but I don't believe there are conclusive arguments for or against 1). I believe there are very strong arguments against 2).

[evan]

^my argument goes more like this:

p) Everything is made up of functions (only)
p) We are made up of functions
C1) Therefore, everything we do can be ascribed to functions
C2) We can define the things we do as functions

p) Computers implement functions
C3) A computer could implement the functions we use (given that they are defined, which is obviously the hard part)
C4) That computer would literally fit the definitions we have for certain things we do, such as "feeling", "thinking", "living" if it implemented those functions because of C2.


"Everything is functions" is a basic assumption of science. The point of science is to formally define phenomena in terms of functions. The better science gets, the better the definitions get. We may never entirely break down humans into functions, but as we get closer, we'll get closer and closer to building computers that we will ascribe human characteristics to.

Anyway, if we fully defined "consciousness" as a set of necessary and/or sufficient features, we could build a computer that had those features even without having to copy all the functions that make up humans.

[ygolo]

OK. So you arent saying "Everything is computing?"

There are two main problems with your argument you presented.

1) "Everything is made up of functions" is vague. How, for instance, are quarks "made up of functions?"

2) It is Mathematically provable that there exist mathematically well-defined functions that are not computable.

Here is the proof:

Take any real number a, define f_a(x)=a, for all real-numbers x. f_a is a well defined mathematical function, no matter what the selection of a. This means, since there are uncountably many real numbers, there are uncountably many well-defined functions.

However, by the Church-Turing Thesis each computable function is computed by some Turing machine. A Turing machine can be represented as a finite string of symbols over a finite alphabet of symbols. The set of all finite strings over a finite alphabet is countable, so the set of all Turing machines is countable, which means the set of all computable functions is countable.

This means there exists uncountably many uncomputable well-defined functions.

What if one of those well-defined functions is a function describing the human reasoning process?

A pithy way to summarize the this view-point is to say:
The world (and the mind) are analog, while computation (along with the deductive processes that spawn it) is digital.

If you are going to change the what you are calling "computing" from something equivalent to a Turing machine of some-sort, to something else, that needs to be made clear.

Because your argument is not valid based on what the standard definition of computing is.

(as an interesting side note, look up the Busy Beaver Function as a specific example of an uncomputable but well-defined function, its cool).

[evan]

Originally Posted by ygolo

OK. So you arent saying "Everything is computing?"

There are two main problems with your argument you presented.

1) "Everything is made up of functions" is vague. How, for instance, are quarks "made up of functions?"

Eh. I dunno really. Just like I don't know how humans are made up of functions. It's just a framework, not that there aren't other viable frameworks.

2) It is Mathematically provable that there exist mathematically well-defined functions that are not computable.

Hmm. Yeah maybe my wording doesn't exactly work. I guess I'd just say that the uncomputable functions (of which there are uncountably many, like you said) just aren't implementable. So then why would you think humans implement them?

I'm not saying I can prove that humans don't implement uncomputable functions, I'm just saying it seems like a more complicated philosophical worldview to have, and I'm using Occam's Razor.

[ygolo]

Originally Posted by dissonance

I'm not saying I can prove that humans don't implement uncomputable functions, I'm just saying it seems like a more complicated philosophical worldview to have, and I'm using Occam's Razor.

I lean towards agreeing with you, but this belief for me is more like believing a slightly better team will win a ball-game.

Besides, there really no reason to pick a side on this matter right now.

It would be a more complex worldview to believe than humans implement uncomputable functions, but for those who believe that there is a God, and that our minds are reflections of God's mind, or something along those lines, it would be quite natural.

There are also many, scientists included, that believe that consciousness occurs at a scale below and speed above what the uncertainty principle allows us to measure.