Any science of logic?

What is the basis for reason? And mathematics?

Moderators: AMod, iMod

Univalence
Posts: 492
Joined: Sun May 12, 2019 6:28 pm

Re: Any science of logic?

Post by Univalence »

Speakpigeon wrote: Fri May 17, 2019 11:28 am And I've looked at what the so-called experts talk about and I've found what they say lacking.
So is it a case of 'I will know it when I see it' then?

If you are finding the work of experts lacking, then it seems to me you have a much broader perspective than the 'experts' do.
If that's the case why are you waiting for somebody else to write the book?

Why don't you finish Boole's work?
Speakpigeon wrote: Fri May 17, 2019 11:28 am No, I mean human logic.
It is obvious to me that the expression 'human logic' means something very different to you than it does to me.

Is it not obvious to you?
Speakpigeon wrote: Fri May 17, 2019 11:28 am What's even difficult to understand? How could you possibly go through life without understanding what such a simple expression as "human logic" means?
Daniel Dennett has a phrase which captures all I need to say: Competence without comprehension.

There is a difference between using something and understanding it.
You can USE a car without understanding how it works.
You are busy USING a computer without understanding how it works.

Similarly: you can USE your mind without UNDERSTANDING it.

I'd venture a guess that the verb "to understand" means something very different to you than it means to me.
Speakpigeon wrote: Fri May 17, 2019 11:28 am The logic of the human mind? The logic of rational thinking? Do you want me to explain what I mean by human being?
And, you don't even seem to understand the word "since". :roll:
My understanding of 'human logic' is approximately the body of knowledge behind 'computational trinitarianism'.
I understand 'human logic' as the science of constructing and executing models of reality for the purpose of computing consequences.
I understand 'rational thought' as instrumental thought-patterns used by humans towards achieving a particular, subjective goal.

Rationality is goal-optimisation.
Attempting to 'accurately describe human logic' is one particular kind of goal.

It brings us no closer to having objective criteria for success or failure.

I do not find a computational description of my mind lacking at all. Broadly speaking - it is how my mind works.

If you disagree with my conception of 'human logic' and you are finding it lacking then tell us why.
Would you say that it's outright wrong, or just incomplete?
Speakpigeon wrote: Fri May 17, 2019 11:28 am Sorry, I don't seem to see the end of your sentence here... What's the question?
When I look at the work of all Mathematicians, Logicians and Computer scientists from a holistic (Quinean?) lens I have arrived at the 'computational trinitarianism' perspective. It is a phenomenological perspective - computation is a phenomenon. I also find it to be an accurate-but-incomplete description of my decision-making mind.

Why do you find it lacking?
Speakpigeon wrote: Fri May 17, 2019 11:28 am Mathematicians want to be the best among their peers.
I don't know if I can label myself a mathematician, but I DO mathematics. Often. Daily perhaps?

The above sentence is not representative of my desires.

What I want out of computation is the ability to outsource the tedious, boring and repetitive mental work to a mechanical mind.
Principle of least effort and all that.

I want to automate myself away, so that I can sit on a beach drinking cocktails while machines do the undifferentiated heavy lifting of human existence.
Speakpigeon wrote: Fri May 17, 2019 11:28 am Think of games. People want to play, whatever the game. But you don't get to play if you insist on redefining the rules. Once in a while you'll have individuals who decided to upend the table, Copernicus, Kepler, Einstein. For any Einstein, you could quote thousands of experts who asserted that Newton was correct. In fact, they're still saying it today!
I mean I'm sure these people were bright minds but so what? You think a bright mind is necessarily correct?
Basically, these people didn't even look, really, at human logic. That's all there is to it.
Perhaps you have heard of Game semantics.

In the same way that I wouldn't label myself a mathematician, I wouldn't label myself a game theorist either.

But I DO game theory every single day of my life.

I play the game of life. I develop strategies towards achieving my goals knowing that reality is subject to chance and stochastics.
To this end I use many of the tools of logic. In particular probability theory and I use computers to perform Monte Carlo simulations.

I use the results of these logics to make decisions.
Speakpigeon wrote: Fri May 17, 2019 11:28 am If something say the Moon is square and green, I infer they're not talking about what I call the Moon.
Good! Now apply that same thought-pattern back to 'human logic'!

When you say to me 'human logic' is not computational, then you are not talking about what I call 'human logic'.
Speakpigeon wrote: Fri May 17, 2019 11:28 am Aristotle's syllogistic would have to be a start.
So that constrains the discussion only to deduction.

And what about induction?
What about counter-factual reasoning?
What about pattern recognition?
What about judgments?

Your mind doesn't do any of those things?

Most concerning of all. Deductive logic is entirely deterministic, but the world we live in is stochastic.

It sure seems like the wrong tool for the job.

Speakpigeon wrote: Fri May 17, 2019 11:28 am I can't vouch for all of them being inappropriate but each one I took the time to look at is indeed inappropriate and I think I looked at the main ones. I thought initially Gentzen's natural deduction was likely correct but turns out it isn't.
Well, there are no esaping the epistemic problem of criterion here.

Could you verbalize your notions of 'correctness' and 'incorrectness' when it comes to 'accurate descriptions of human logic'.

What is your referent for 'correctness'?
Speakpigeon wrote: Wed May 15, 2019 4:53 pm I can only use my own intuition. That's all each of us has. Aristotle's syllogistic is faultless. A bit limited in scope but all good. The Stoics seem all good to me but I haven't looked at the details. I'm fine with all widely accepted logical truths since Aristotle. Isn't that a solid foundation? Well, not quite but that's a start. Most logicians nowadays poo-poo intuition, essentially because respected mathematicians in the past have been guilty of committing themselves to quite a few paradoxes. I disagree. We only have our intuition just as we only have our senses of perception. Logical intuition is essentially like a sense of perception. Too me, it is exactly that.
Is it possible that your intuition works differently to my intuition?
Is it possible that all humans don't think the same way?
Is it possible to develop new intuitions and discard old intuitions?

Is it possible that your mind works differently to my mind?

Which would explain why 'computational trinitarianism' may seem like an accurate description of 'human logic' to me, while it doesn't appear so to you.
Speakpigeon wrote: Wed May 15, 2019 4:53 pm Yes, I am. Were not sheep, you know.
I didn't say that I am a sheep. But I am a computer. Not in the ontological, but in the metaphysical sense.

I spend sufficient amount of my life designing/conceptualising/creating things. I use my mind in a highly collaborative space and so I am finding myself needing a nomenclature/concepts/models to talk about my mind. To this goal I use the models of computer science and It works. Which is all I could ever ask from a scientific model.

If you give me something that works better - I am happy to adopt it.

Hence why I keep repeating: Computational trinitarianism is an accurate-but-incomplete description of my mind.
Speakpigeon wrote: Wed May 15, 2019 4:53 pm Quote somebody with the stature of a Boole, Frege, Russell, Tarsky, Quine etc.
That's a nice attempt at dodging the point. You are still appealing to the authority of (dead) philosophers.
It's ironic that none of the authorities you recognise ever worked on computation, so I am not sure why I would ever quote them on matters of 'computational trinitarianism'.

You are welcome to wait until Scott Aronson dies before you take him seriously.
Or you can read Marvin Minsky's views
in his book 'Society of Mind" he describes the human mind as a distributed system

Of course, no matter how many 'experts' I quote, it still doesn't change the fact that you are insisting I must appeal to an authority in order to support my views. That seems very.... irrational.

Science is the belief in the ignorance of experts. --Richard Feynman

If the evidence alone doesn't convince you and you still insist on settling this via an authority then here is a list of authors who have written on the Computational Theory of Mind.
Speakpigeon wrote: Wed May 15, 2019 4:53 pm So designing a new fridge or a new toaster is doing something like inventing a new kind of human logic?!
This is a bad analogy.

If you have never designed a toaster before, but you want to know how toasters work you are going to study them first.

Once you have sufficiently understood the inner workings of a toaster you are going to have to generate some jargon to describe what you have seen, experienced and learned. You are going to need words like 'timer', 'timer delay mechanism', 'heating element', 'bread holder' and 'crumb collector'.

If you have sufficiently understood how toasters work you ought to be able to make one from scratch that works much like your referent.
If you can't make a 2nd toaster that works just like your referent, can you really say that you 'understand' how toasters work?

Or you could recognize that the essence of a toaster is its purpose, and you could come up with a new and innovative machine that revolutionises the way we currently make toast.

As the elderly say: there are many ways to skin a cat.
Speakpigeon wrote: Wed May 15, 2019 4:53 pm OK, you believe there's no logic that's intrinsic to the human brain. Me, I do. And I believe that if we didn't have that, we couldn't design any new toaster, let alone some computer. In fact, I believe we couldn't even make sense of the world around us.
It's not that. I just think ontological questions are a dead end.

Science doesn't explain what things ARE.
Science explain what things DO and how things WORK.
Science often produces a number of competing models, that explain different aspects of the same thing.

And so (given the title of this thread) my epistemic criterion for 'accurate description of the human mind' stops at functional equivalence.

If I can build a machine that works LIKE a human mind, then I can say that I understand how the mind works.
For better or worse - self driving cars are a manifestation of human logic in a machine.
User avatar
Speakpigeon
Posts: 987
Joined: Sat Nov 11, 2017 3:20 pm
Location: Paris, France, EU

Re: Any science of logic?

Post by Speakpigeon »

Univalence wrote: Fri May 17, 2019 12:49 pm Well, there are no esaping the epistemic problem of criterion here. Could you verbalize your notions of 'correctness' and 'incorrectness' when it comes to 'accurate descriptions of human logic'. What is your referent for 'correctness'?
I already told you. The standard is the human mind.
Univalence wrote: Fri May 17, 2019 12:49 pm
Speakpigeon wrote: Wed May 15, 2019 4:53 pm We only have our intuition just as we only have our senses of perception. Logical intuition is essentially like a sense of perception. Too me, it is exactly that.
Is it possible that your intuition works differently to my intuition?
Is it possible that all humans don't think the same way?
Is it possible to develop new intuitions and discard old intuitions?
Is it possible that your mind works differently to my mind?
I'm sure there are differences. However, I believe there is such a thing as the human brain's logical capability. Our logical intuition is a by-product of that. Differences between people are not fundamental. They are of the same sort as the differences between the design on US tanks and Russian tanks. Same logic, but differences in the engineering culture, economic constraints, availability of base materials and what not. But the logic is the same. Different inputs, therefore different outputs. Tell me if you disagree with the modus ponens... No? The modus tollens perhaps? The hypothetical syllogism, then? Transposition? No?
Univalence wrote: Fri May 17, 2019 12:49 pm
Speakpigeon wrote: Wed May 15, 2019 4:53 pm Quote somebody with the stature of a Boole, Frege, Russell, Tarsky, Quine etc.
That's a nice attempt at dodging the point. You are still appealing to the authority of (dead) philosophers.
I didn't ask you to quote them. I asked for someone of their calibre and active now.
Univalence wrote: Fri May 17, 2019 12:49 pm Of course, no matter how many 'experts' I quote, it still doesn't change the fact that you are insisting I must appeal to an authority in order to support my views. That seems very.... irrational.
You still missing the point. I'm trying to see if any mathematician today tries to model human logic as Boole and Frege thought they were doing.
Univalence wrote: Fri May 17, 2019 12:49 pm And so (given the title of this thread) my epistemic criterion for 'accurate description of the human mind' stops at functional equivalence.
If I can build a machine that works LIKE a human mind, then I can say that I understand how the mind works. For better or worse - self driving cars are a manifestation of human logic in a machine.
Functional equivalence has to be good enough. I'm not looking for metaphysical answers. Your claim, however, is without foundation. It's based on a principle. I'm looking for serious attempt to emulate human logic, even flawed ones. To hope to achieve that, though, you would need to understand human logic. Just claiming the brain is a computer and proceed with work on computers won't do. The reference is the human mind. There is no other.
EB
Univalence
Posts: 492
Joined: Sun May 12, 2019 6:28 pm

Re: Any science of logic?

Post by Univalence »

Speakpigeon wrote: Fri May 17, 2019 5:40 pm I already told you. The standard is the human mind.
Ok, but "the mind" is not an ontological entity of any kind. It's an emergent phenomenon.
Minds are what brains DO.

So I am asking you to focus on the ontology of a "description" and not on the ontology of a "human mind".
What IS an "accurate description"?

Suppose I gave you a paper titled "The empirically verified, peer-reviewed model of the human mind" which contains 1000 pages of mathematical equations. All scientists agree - it is correct. Great! What does that mean to you when you can't even read the Mathematics?

Beyond accepting the authority of the institution of science, how would YOU go about asserting that the mathematical model accurately describes the human mind?

What is your, subjective criterion for "model accuracy" if your referent is the very thing you are attempting to model?
Speakpigeon wrote: Fri May 17, 2019 5:40 pm I'm sure there are differences. However, I believe there is such a thing as the human brain's logical capability.
That's a vague truism. If the human mind can invent logics then it clearly has a capability for logic.
Speakpigeon wrote: Fri May 17, 2019 5:40 pm Our logical intuition is a by-product of that. Differences between people are not fundamental. They are of the same sort as the differences between the design on US tanks and Russian tanks. Same logic, but differences in the engineering culture, economic constraints, availability of base materials and what not. But the logic is the same.
Another bad analogy. It was you who claimed that the mind is not designed. Tanks are.
The "logic" behind a tank is teleological. The "logic" of a tank is the purpose for which it was built.

Do minds have a teleology?
Speakpigeon wrote: Fri May 17, 2019 5:40 pm Different inputs, therefore different outputs.
Inputs and outputs sounds a lot like the language of computation.
It sounds a lot like a black box.

Speakpigeon wrote: Fri May 17, 2019 5:40 pm Tell me if you disagree with the modus ponens... No? The modus tollens perhaps? The hypothetical syllogism, then? Transposition? No?
I neither agree nor disagree with them. I see them for what they are - rules in a game.

If you are playing the game of "propositional logic" transposition is a valid rule of inference. If you are playing the game of "intuitionistic logic" it is not.

https://proofwiki.org/wiki/Rule_of_Tran ... tion/Proof
This theorem depends on the Law of the Excluded Middle, by way of Double Negation Elimination.
This is one of the axioms of logic that was determined by Aristotle, and forms part of the backbone of classical (Aristotelian) logic.
However, the intuitionist school rejects the Law of the Excluded Middle as a valid logical axiom. This in turn invalidates this theorem from an intuitionistic perspective.
When they are useful - I use them.
When they are not useful - I don't use them.

It really brings us back to teleology: why would you want to play the game of "classical logic", or the game of "propositional logic", or "intuitionistic logic". What is the purpose of playing such games?
Speakpigeon wrote: Fri May 17, 2019 5:40 pm You still missing the point. I'm trying to see if any mathematician today tries to model human logic as Boole and Frege thought they were doing.
We clearly come from very different philosophical foundations here. The very process of modeling is human logic IN ACTION.

The EXPRESSION of human logic. The uttering of symbols like "A ∧ ¬B" is grammar. It's just language. A mechanism for the mind's self-expression.
Certain grammars allow for expressing certain kinds of phenomena.

Temporal logic models temporal phenomena.
Modal logic models modal phenomena.
Epistemic logic models epistemic phenomena.

The particular instance where the human mind attempts to model itself is interesting only because we have an entity trying to model itself.
It is interesting because it's self-referential/recursive. Recursion grammars are computation
Speakpigeon wrote: Fri May 17, 2019 5:40 pm Functional equivalence has to be good enough.
OK, then lests stick to functional equivalence.
Speakpigeon wrote: Fri May 17, 2019 5:40 pm Your claim, however, is without foundation. It's based on a principle. I'm looking for serious attempt to emulate human logic, even flawed ones.
In 2019 we have self-driving cars, self-flying airplanes. Facial and voice recognition systems. Machine learning algorithms which excel at tasks which used to be strictly the domain of human ingenuity. Like playing Chess and Go.

We have all sorts of automation that DOES some parts of what human minds DO.

All of this emulation is written in programming languages. All of this emulation runs on a computers.

You don't see those are successful attempts at emulating the human mind?
Speakpigeon wrote: Fri May 17, 2019 5:40 pm To hope to achieve that, though, you would need to understand human logic.
You have claimed that you understand human logic (by virtue of being unable to function in life if you don't).
So if you understand human logic - why aren't you describing it?
Speakpigeon wrote: Fri May 17, 2019 5:40 pm Just claiming the brain is a computer and proceed with work on computers won't do. The reference is the human mind. There is no other.
You are not understanding me. I did not claim that the brain IS a computer. I do not make ontological claims of any kind.
I claimed that the brain DOES computation.

And we have already have mathematical models for computation.
Last edited by Univalence on Sat May 18, 2019 12:24 pm, edited 2 times in total.
Univalence
Posts: 492
Joined: Sun May 12, 2019 6:28 pm

Re: Any science of logic?

Post by Univalence »

This back-and-forth is getting us further away from the most pertinent question.
Speakpigeon wrote: Fri May 17, 2019 11:28 am How could you possibly go through life without understanding what such a simple expression as "human logic" means?
You are going through life. By your very own modus tollens you are necessarily implying that you understand what "human logic" means.

Can you describe it to us?
User avatar
Speakpigeon
Posts: 987
Joined: Sat Nov 11, 2017 3:20 pm
Location: Paris, France, EU

Re: Any science of logic?

Post by Speakpigeon »

Univalence wrote: Sat May 18, 2019 10:40 am
Speakpigeon wrote: Fri May 17, 2019 5:40 pm I already told you. The standard is the human mind.
Ok, but "the mind" is not an ontological entity of any kind. It's an emergent phenomenon.
Minds are what brains DO.

Absolutely, but brains aren't any ontological entity either, certainly not as far as anyone seems to know.
Still, for all practical purposes, we can observe what our own brain does and that's called the mind. That will have to do up until such a time as science tells us what it is our own brain does. And I don't feel like it would be a good idea to wait to that.
Univalence wrote: Sat May 18, 2019 10:40 am So I am asking you to focus on the ontology of a "description" and not on the ontology of a "human mind".
???
What's the "ontology of a description"?!
And why are you so keen on the word "ontology"? You also say "functional equivalence"... Let's do that instead.
Descriptions are good enough as they are. No need for any ontologically searching questions.
Univalence wrote: Sat May 18, 2019 10:40 am What IS an "accurate description"?
I have no idea. I can tell whenever a description is appropriate, but even then I don't think I could exclude there's a better description.
Univalence wrote: Sat May 18, 2019 10:40 am Suppose I gave you a paper titled "The empirically verified, peer-reviewed model of the human mind" which contains 1000 pages of mathematical equations. All scientists agree - it is correct. Great! What does that mean to you when you can't even read the Mathematics?
That's why I accept the idea of functional equivalence. If I don't understand the mathematics, it's not equivalent for me. If any scientist wanted to convince me, then I'm afraid he would have to speak in plain French or plain English. Or I could learn the mathematics. I'll do whatever seems the best option.
Univalence wrote: Sat May 18, 2019 10:40 am Beyond accepting the authority of the institution of science, how would YOU go about asserting that the mathematical model accurately describes the human mind?
I would look at the predictions. That what functional equivalence means.
Univalence wrote: Sat May 18, 2019 10:40 am What is your, subjective criterion for "model accuracy" if your referent is the very thing you are attempting to model?
Sorry, I can't parse "subjective criterion"... The word "criterion" is good enough, you know?
I'm not attempting to model my mind. I'm taking some aspects of my mind, what I call my logical intuition, to be an accurate indication of the logic of my brain. I also look at what people have said about logic, Aristotle, the Stoics, the Scholastics, even mathematicians today but to a lesser extent. So, in effect, I have several criteria but I accept my intuition as the main criterion. But another criterion is to prove all the logical truths as identified by the tradition unless I could understand why it wouldn't be in fact a logical truth. So, transposition may not be a logical truth, but the reason given by intuitionists doesn't convince me.
EB
User avatar
Speakpigeon
Posts: 987
Joined: Sat Nov 11, 2017 3:20 pm
Location: Paris, France, EU

Re: Any science of logic?

Post by Speakpigeon »

Univalence wrote: Sat May 18, 2019 10:40 am
Speakpigeon wrote: Fri May 17, 2019 5:40 pm You still missing the point. I'm trying to see if any mathematician today tries to model human logic as Boole and Frege thought they were doing.
We clearly come from very different philosophical foundations here. The very process of modeling is human logic IN ACTION.

The EXPRESSION of human logic. The uttering of symbols like "A ∧ ¬B" is grammar. It's just language. A mechanism for the mind's self-expression.
Certain grammars allow for expressing certain kinds of phenomena.

Temporal logic models temporal phenomena.
Modal logic models modal phenomena.
Epistemic logic models epistemic phenomena.
If I had intuitions about temporal logic, maybe i would buy your point but I don't. As I see, there's no doubt as to there being some logical truths. That's my brain saying this through my logical intuition.
Temporal logic, epistemic logic etc. all depend on those basic rules and as such they are not logic proper.
You couldn't even articulate anything about your so-called "temporal logic" if you couldn't rely on your actual sense of logic. Everything you argue here relies on that and not on anything else. All we need to argue in a rational way is logic and the definitions of the words we use to identify what we're arguing about.
You're trying to drown the fish. Logic doesn't disappear just because you want to call "logic" things that are certainly logical but aren't logic as such. The whole of mathematics is logical, and we don't call it "logic". So, as I see it, there's an equivocation on the word "logic" between the logic of the human mind and the logic of toasters. Temporal logic is in the same ballpark as the whole of mathematics as well as toasters. You do that, I'm not too interested. It's legitimate to talk of the logic of a toaster, but you need to keep in mind it's not the same sense of the word "logic":
Logic
1. The study of principles of reasoning, especially of the structure of propositions as distinguished from their content, and of method and validity in deductive reasoning.
2.
a. A system of reasoning: Aristotle's logic.
b. A mode of reasoning: By that logic, we should sell the company tomorrow.
c. The formal, guiding principles of a discipline, school, or science.
3. Valid reasoning: Your paper lacks the logic to prove your thesis.
4. The relationship between elements and between an element and the whole in a set of objects, individuals, principles, or events: There's a certain logic to the motion of rush-hour traffic.
5. Computers
a. The non-arithmetic operations performed by a computer, such as sorting, comparing, and matching, that involve yes-no decisions.
b. Computer circuitry.
c. Graphic representation of computer circuitry.
Me, I'm only interested in sense 3, "valid reasoning".
I call "human logic" whatever capacity of our brain allows us to reason validly.
EB
User avatar
Speakpigeon
Posts: 987
Joined: Sat Nov 11, 2017 3:20 pm
Location: Paris, France, EU

Re: Any science of logic?

Post by Speakpigeon »

Univalence wrote: Sat May 18, 2019 10:40 am
Speakpigeon wrote: Fri May 17, 2019 5:40 pm Your claim, however, is without foundation. It's based on a principle. I'm looking for serious attempt to emulate human logic, even flawed ones.
In 2019 we have self-driving cars, self-flying airplanes. Facial and voice recognition systems. Machine learning algorithms which excel at tasks which used to be strictly the domain of human ingenuity. Like playing Chess and Go.

We have all sorts of automation that DOES some parts of what human minds DO.

All of this emulation is written in programming languages. All of this emulation runs on a computers.

You don't see those are successful attempts at emulating the human mind?
No.
All these things rely on the little bit of formal logic we know, but they're not efforts to model human logic as such. These things are application of human logic. They are uses of human logic, not models of it.
We didn't even have to wait to have any kind of formal logic to start doing this. Pythagoras and Euclid didn't need Aristotle's syllogistic. Any idiot who argue even stupid things relies on logic for doing it and he's not trying to model human logic. He's just using it.
To be honest, I thing the distinction I make is pretty obvious. Any scientist would understand. To model something, you have to observe it and work from the empirical evidence available to you. And that's no longer what mathematical logicians seems to be doing nowadays.
EB
User avatar
Speakpigeon
Posts: 987
Joined: Sat Nov 11, 2017 3:20 pm
Location: Paris, France, EU

Re: Any science of logic?

Post by Speakpigeon »

Univalence wrote: Sat May 18, 2019 10:40 am You are not understanding me. I did not claim that the brain IS a computer. I do not make ontological claims of any kind.
I claimed that the brain DOES computation.
Sure, that's your claim. It's a principle and not empirically verified. I don't think it's even anything like a scientific claim, i.e. something that's falsifiable. How do you think I could possibly prove you claim is wrong?
And I already told you. Quantum physics is more fundamental than the Turing machine and it's the scientific theory. But doing QM doesn't help with human logic, does it?
EB
wtf
Posts: 1179
Joined: Tue Sep 08, 2015 11:36 pm

Re: Any science of logic?

Post by wtf »

Univalence wrote: Sat May 18, 2019 10:40 am You are not understanding me. I did not claim that the brain IS a computer. I do not make ontological claims of any kind.
I claimed that the brain DOES computation.
Ooh! I'll take the other side of that one.

First, a semantic point. You say a brain does computation, but that it's not a computer. I'm not sure what that means. My clothes dryer dries clothes. A machine that DOES clothes drying IS a clothes dryer. A machine that DOES computation IS a computer.

Unless you mean that the brain CAN do computation but that it ALSO does more. That I would agree with. Can you clarify? For example a combined washer dryer DOES dry clothes but it also washes them, so we can't call it a dryer. Is that what you mean?

If you are saying the brain is ONLY a Turing machine, I'll gladly debate that.

But if you are saying that the brain CAN compute, as when we use pencil and paper to execute the Euclidean algorithm; but that brains do something MORE than mere computation in general; then we have no difference of opinion.

Please clarify. Thanks.
Univalence
Posts: 492
Joined: Sun May 12, 2019 6:28 pm

Re: Any science of logic?

Post by Univalence »

wtf wrote: Sun May 19, 2019 9:52 pm First, a semantic point. You say a brain does computation, but that it's not a computer. I'm not sure what that means.
My intention behind saying that was to establish if I am talking to an ontologist. Somebody who is seeking to discover the "true nature of the mind".
We have moved on from that and we have agreed that functional equivalence is a sufficient criterion.

But we aren't out of philosophical waters yet, because it boils down to a decision problem.
Given two machines who decides whether they are "functionally equivalent" and how?

The thought experiment here is the Turing test, but the standard flaws with the verifications perspective apply to the Turing test also.
Your ability to discern the difference between two things stops at the fidelity of your own understanding.
wtf wrote: Sun May 19, 2019 9:52 pm My clothes dryer dries clothes. A machine that DOES clothes drying IS a clothes dryer. A machine that DOES computation IS a computer.
Your room heater dries clothes also.
Your oven dries clothes also.
Your table, when placed in direct sunlight, dries clothes also.

Coming from an instrumentalist view-point, the functional equivalence of any two particular machines cannot be asserted without criteria for success.
And criteria for success cannot be asserted without a clear goal in mind.

If your goal is to dry your clotes then all of the above machines are functionally equivalent.
If your goal is to cook dinner - they aren't.

So I think it's far more accurate to say: anything which can be USED FOR drying clothes can be a clothes dryer.
Similarly. Anything that can be USED FOR computation can be a computer.

And a whole lot of things can be used for computation apparently.

Mechanical gears
Atoms
Water

The last link above makes a rather bold claim. Computation is the manipulation of information (and is bound by the laws of physics). I quite like that description.
wtf wrote: Sun May 19, 2019 9:52 pm Unless you mean that the brain CAN do computation but that it ALSO does more. That I would agree with.
If you are saying the brain is ONLY a Turing machine, I'll gladly debate that.
This kind of argumentation is not going to get us anywhere unless you are explicit about the things which fall under your category of "MORE" that you think are non-computational.

To me this is magical thikning. Down the layers of abstraction a brain is made up of quarks, leptons and bosons (like everything in this damn world).
In 2019 we can compute the consequences of chemical reactions from the laws of QM.
And we have successfully digitised the nervous system of a roundworm.

I mean it's plausible that we, humans, are special, but I have no reason to believe that. We are just very, very complex. So we can say that our brains do things we don't understand, but IF we understood what it is that our brains do then I see no obstacles to reproducing the functionality.
wtf wrote: Sun May 19, 2019 9:52 pm But if you are saying that the brain CAN compute, as when we use pencil and paper to execute the Euclidean algorithm; but that brains do something MORE than mere computation in general; then we have no difference of opinion.
I think you have a rather narrow view of the field. It takes a little bit more than executing Euclid's algorithm to drive a car on public roads.

I am talking about inter-connected, modular functionality/specialization, distributed systems, control theory, complexity theory, adaptive systems, self-healing systems, machine learning, computer vision/hearing, natural language processing, hypothesis/plausibility testing, decision-making under uncertainty.

I really think it's much easier to focus on the functions which you deem non-computational. I'll give you one... autonomous goal-setting. Wants/desires.
wtf
Posts: 1179
Joined: Tue Sep 08, 2015 11:36 pm

Re: Any science of logic?

Post by wtf »

Univalence wrote: Mon May 20, 2019 9:03 pm I really think it's much easier to focus on the functions which you deem non-computational. I'll give you one... autonomous goal-setting. Wants/desires.
As usual, a pile of word salad instead of a direct response to my simple questions.

Someone privately pointed out to me who you are. I'm glad you're not dead, just a sockpuppet.
Univalence
Posts: 492
Joined: Sun May 12, 2019 6:28 pm

Re: Any science of logic?

Post by Univalence »

wtf wrote: Mon May 20, 2019 9:16 pm As usual, a pile of word salad instead of a direct response to my simple questions.
You mean your over-simplified question...

If we are talking about functional equivalence (and since there's a clear trend in the direction of successfully automating minds) what would you deem a function that is impossible to replicate?

P.S It sounds like "word salad" because we are tripping over this phenomenon.
wtf
Posts: 1179
Joined: Tue Sep 08, 2015 11:36 pm

Re: Any science of logic?

Post by wtf »

Univalence wrote: Mon May 20, 2019 9:18 pm
wtf wrote: Mon May 20, 2019 9:16 pm As usual, a pile of word salad instead of a direct response to my simple questions.
You mean your over-simplified question...
Well then answer it for my over-simplified mind. Perhaps your wordy and convoluted posts are a manifestation of your wordy and convoluted thinking.
Univalence
Posts: 492
Joined: Sun May 12, 2019 6:28 pm

Re: Any science of logic?

Post by Univalence »

wtf wrote: Mon May 20, 2019 9:45 pm Well then answer it for my over-simplified mind.
OK. 42.

I gave you 3 examples of things which dry clothes which are not clothes dryers.
If you care to elaborate on what you are asking - maybe I can be more precise in my response.

In particular, you could elucidate what you mean by "it also does more", because I also gave you a bunch of examples where computation is used for significantly more complex real-world tasks than your grade 3 maths homework.
wtf wrote: Mon May 20, 2019 9:45 pm Perhaps your wordy and convoluted posts are a manifestation of your wordy and convoluted thinking.
The complexity of my thought corresponds to the problems I work on. Perhaps you expect simple answers to complex problems?
There's a reason why we call them complex systems.

Perhaps your maths background has biased your thinking towards monoliths, symmetry and homogeny? Large-scale distributed systems are nothing like that. They are modular, asymmetrical and heterogenous. They behave like organisms. And they have a whole lot of non-trivial side-effects and emergent behaviours, which would offend any mathematician's notions of purity and beauty.

But they work.
wtf
Posts: 1179
Joined: Tue Sep 08, 2015 11:36 pm

Re: Any science of logic?

Post by wtf »

Univalence wrote: Mon May 20, 2019 9:54 pm I gave you 3 examples of things which dry clothes which are not clothes dryers.
If you care to elaborate on what you are asking - maybe I can be more precise in my response.
You said brains compute but are not computers. What exactly do you mean by that? If brains are not computers but they compute, then they must do something IN ADDITION to computing. Is that your point?
Post Reply