Physical Laws - Best Case for ID?

The Incompleteness Theorem poses no problems whatsoever to the idea that minds are purely physical in origin, because we have no evidence whatsoever that minds are not subject to the Incompleteness Theorem.

Hey, Pixy. Would you mind giving me the rundown on what is meant by "completeness" and formal systems ? Having a hard time following.
 
Hey, Pixy. Would you mind giving me the rundown on what is meant by "completeness" and formal systems ? Having a hard time following.

There's a nice little description of completeness and consistency here (ignore everything after the first few paragraphs, though). I was going to link to wikipedia, but it seems to be having trouble at the moment :S

A formal system has a certain associated formal language, and a set of associated formal axioms and rules of inference. "Formal" here means that the language and the axioms are defined precisely, like the syntax of a programming language, in a way that makes it a mechanical matter to decide whether something is a proof in the system. When we say that a formal system is complete, this also has a precise sense: for every sentence A in the language of the system, either A or its negation is a theorem of the system, i.e. can be derived using the axioms and rules of inference. Similarly, consistency has a precise sense: the system is consistent if there is no A such that both A and its negation can be derived.

Anyway, Godel's Theorem states that any reasonably complex formal system can never be complete and consistent: you can always construct propositions that can not be proved neither to be true or false. It's a very interesting result, and basically p**ses all over the idea that you could build a perfect formal system (which was always the goal of mathematics)

I've never seen why Godel's Theorem provides such a difficulty for the possibility of Machine Intelligence. Why would a machine intelligence need to be either complete and/or consistent when we so obviously are not? Also, since modern AI tends to deal with probabilities rather than absolutes, I fail to see how it applies at all.
 
Completeness and Incompleteness and all that

Imagine the set of all possible statements.

Think of a logical system (formal system) as a set of statements that are assumed to be true (axioms) and logical machinery to derive new statements from existing statements (e.g. logical deduction).

We need to assume that the axioms are consistent - that they are free of contradiction - or the whole thing falls down like a house of cards.

Use a logic rule to produce a new statement from the axioms. This new statement is said to be 'derived' from the axioms and we can point to the rule and the statements that were fed to it to produce the derived statement.

Now use the derived statement and perhaps some of the axioms as input to a logic rule. A new derived statement will be produced. It is also derived from the axioms (although indirectly).

If we can construct a chain of derivations that start with the axioms and end with a statement, then that statement is said to be derivable.

Every statement in the set of all possible statements is either derivable from the set of axioms or it is not.

Statements are also either true or false with respect to the set of axioms.

You can call the set of axioms a 'theory' and all of the statements that are true with respect to the set of axioms the 'theorems' of the theory.

If every derivable statement is true, then the logical system is said to be 'sound'.

If every true statement is derivable, then the logical system is said to be 'complete'.

If some true statement is not derivable, then the logical system is said to be 'incomplete'.

Since the logical mechanism (e.g. deduction) is how we prove something, an incomplete system has statements that are true that can not be proven to be true.

The first-order predicate calculus and the second-order predicate calculus are two logical systems. The second-order predicate calculus contains all of the first-order-predicate calculus as well as the ability to represent and use the concepts of 'there exists' and 'for every'. The second-order predicate calculus is more powerful than the first-order predicate calculus.

The second-order predicate calculus is the system needed for much of mathematics. [not quite true, but would be a real derailer]

Kurt Gödel (a contemporary and collaborator of Einstein) proved that the first-order predicate calculus is complete - that every true statement in the system is derivable (provable).

Kurt Gödel also proved that the second-order predicate is incomplete - that there are true statements in the system that are not derivable (provable).

What Kurt Gödel really proved is that any sufficiently powerful logical system will contain true statements that are not derivable (provable) within that system, and that the second-order predicate calculus is an example of such a system.

Gödel's proof of the incompleteness of second-order predicate calculus stunned the mathematical world, for it said that there will always be mathematical theorems (statements that are true with respect to the axioms) that are not provable.

The consequences of Gödel's (second) proof really haven't spread far outside of the mathematical community. They should - they constitute a revolution.

Hope this helps.
 
Last edited:
...(good description)...

The consequences of Gödel's (second) proof really haven't spread far outside of the mathematical community. They should - they constitute a revolution.

Hope this helps.

Yup, that was much better than my version :)
 
Gödel's proof of the incompleteness of second-order predicate calculus stunned the mathematical world, for it said that there will always be mathematical theorems (statements that are true with respect to the axioms) that are not provable.

The consequences of Gödel's (second) proof really haven't spread far outside of the mathematical community. They should - they constitute a revolution.

I disagree entirely. The consequences of Godel's second proof have already spread far too far, primarily by credophiles (e.g., Penrose) who believe that they constitute a revolution. It's rather like "quantum" in that regard -- people who don't understand the science and the math think that it somehow gives support to their beliefs.

If you really think that the consequences of Godel's second proof should be spread beyond the mathematical community -- what actual consequences does it have? As far as I can tell, the only serious philosophical implications of Godel's proofs are the Kolmogorov-Chaitin definition of information complexity (which is still "mathematics"), Turing's proof that the halting problem is unsolvable (which is still "mathematics" broadly definedand largely of interest only to computing specialists), and the Lucas-Penrose argument about artificial intelligence (which has been proven to be fallacious).
 
I've never seen why Godel's Theorem provides such a difficulty for the possibility of Machine Intelligence. Why would a machine intelligence need to be either complete and/or consistent when we so obviously are not? Also, since modern AI tends to deal with probabilities rather than absolutes, I fail to see how it applies at all.

I agree.

And thanks for the info.
 
I do have more. I think I have some unique ideas, some of which I have had for decades. Every time I saw a new book come out (like Kantor's "information mechanics" 1977), I would get it and look and see if anybody else had thought the same things. Now, of course, with the internet, this is easier to do. I see that some other people have some of the same ideas but I still have a few I have never seen anyone else come up with. I just don't think this forum is the place I want to debut my bottled up ideas.
Probably not. But I do sincerely hope you stop by and fill us in after you've broken your silence. Don't forget to bring your helmet, and watch out for flying bricks. :bricks:
 
The discussion was 'virtual photons' as proposed by Feynman in QED: some go backward in time. I called that 'non-relativistic' since v>c implies 'going backward in time'. I take it you do not like my lexiconical choices.

However you wish to say it, I invited you to explain to me, given a wonderful scientific theory such as QED, where objects such as 'virtual particles' are hypothesized, what do you mean when you say 'they are not real - only models' [my paraphrase since I'm to lazy to go look exactly how you said it]?

If 'virtual photons' are not real, being only a model, what IS really happening? Could there REALLY be little information packets bombarding back and forth between particles, and sometimes arriving before they are sent. If the model uses something that is not real but gives us equisite results, isn't that model the BEST idea, the BEST model we have of what is REAL?
 
Probably not. But I do sincerely hope you stop by and fill us in after you've broken your silence. Don't forget to bring your helmet, and watch out for flying bricks. :bricks:
Probably not? You don't have much confidence in me. I told you that I had the idea about 'horizontal genetic transfer' more than 30 years before it started to become mainstream - or maybe it still isn't what you would call mainstream. I thought it was incredible that persons I told gave me a blank stare when I thought it HAD to be true.

While I probably SHOULD leave, I haven't felt that it is intolerably inhospitable here so you may have to collectively kick me out. I sense some of you are at least sort of curious.
 
Thanks, sphenisc. So, VonNeumann, shall we go with the inter-species vs. intra-species distinction?
I could believe that the African and Asian elephants, while different species, could have descended one from the other. So mere intra-species is not sufficient. Let me come back later with a better definition, but that is semi-close for now.
 
I thought it was incredible that persons I told gave me a blank stare when I thought it HAD to be true.
A blank stare seems pretty appropriate to me. Why do you think it's incredible?

The only reason you are getting answers here is that blank stares don't translate well to web forums.
 
I could believe that the African and Asian elephants, while different species, could have descended one from the other. So mere intra-species is not sufficient. Let me come back later with a better definition, but that is semi-close for now.
I have every confidence you'll come up with something better than "horses is horses", which is essentially the definition I got last time I tried to get someone to nail it down.
 
I have every confidence you'll come up with something better than "horses is horses", which is essentially the definition I got last time I tried to get someone to nail it down.
I have no problem with so-called micro-evolution. I can see that there is already a huge variability in the genome of canines and adaptation within the given variability is pretty obvious. I imagine you could isolate some dogs and within a million years they may not be able to cross breed. If this is micro-evolution then I believe in micro-evolution. But I could also believe that apes could have evolved from humans That would also be micro-evolution. One ingredient of micro-evolution is it results in variability already in the genetic data. Another added ingredient would allow for some translocations or other small scrambling of the information that was already there. Enough said about the RM side of the RM/NS model of micro-e. Now for the NS side. The natural selection side of the model is pretty simplistic, in my thinking, and also there is nothing to argue here. Not if there is already a sufficient amount of biological systems present and competing and all that. I see NS as simple 'negative feedback' that 'deletes' products of RM that are not selected because they are not well adapted .

Macro-e is hard to define. One definition could be 'that which is NOT micro-e' but evolutionists believe macro-e is simply an extrapolation of micro-e. I am astonished at that simplistic thinking because it is not very critical thinking to simply say that a bunch of micro-e over enough time produces macro-e. There are two points I would introduce for my scope of macro-e:
1. macro-e must result in significantly sizeable new information
2. natural selection must stay neutral during the micro-e process, only to become a stabilizer at the end of the macro-e process.

So item#1, that macro produces sizeable increases in the genome, representing new information, is based on my belief that there are no intermediate steps for some structures. Take as an example chlorophyll, even though it may not be a good enough example. If there are no light-to-energy conversion molecules lesser than chlorophyll, there are no precursors. So chlorophyll had to be coded for completely at once. This would be macro evolution unless there was already a similar molecule for some other purpose, that could be changed by a micro-e process. The test for macro-e is that a new sequence of significant size appears that RM would have to work on for too long a period of time in order to get in one fell swoop. It is not all that many nucleotides of a specific sequence that it would take more than 20 billion years to occur by chance. ...Even if the primordial soup filled the whole universe volume. Macro evolution is where 'there's not enough time' because the intermediate steps are too large to be micro-e. Natural selection must remain neutral during the intermediate steps that do not express some advantage. So in macro-e, it would be best if natural selection would look the other way for a while. Natural selection can not do anything but stay neutral to help along the intermediate steps. That is another characterisitic of macro-evolution - being where NS must be neutral while new information is forming.
 
I disagree entirely. The consequences of Godel's second proof have already spread far too far, primarily by credophiles (e.g., Penrose) who believe that they constitute a revolution. It's rather like "quantum" in that regard -- people who don't understand the science and the math think that it somehow gives support to their beliefs.

If you really think that the consequences of Godel's second proof should be spread beyond the mathematical community -- what actual consequences does it have? As far as I can tell, the only serious philosophical implications of Godel's proofs are the Kolmogorov-Chaitin definition of information complexity (which is still "mathematics"), Turing's proof that the halting problem is unsolvable (which is still "mathematics" broadly definedand largely of interest only to computing specialists), and the Lucas-Penrose argument about artificial intelligence (which has been proven to be fallacious).
Could you aim me toward this proof of fallaciousness re Penrose/Lucas?
 
One definition could be 'that which is NOT micro-e' but evolutionists believe macro-e is simply an extrapolation of micro-e. I am astonished at that simplistic thinking because it is not very critical thinking to simply say that a bunch of micro-e over enough time produces macro-e.
It's not simplistic. It's just quantitative. You have small amounts of change. You have large amounts of change. You have medium amounts of change. If you keep making small changes, you may end up with a big change.

There are two points I would introduce for my scope of macro-e:
1. macro-e must result in significantly sizeable new information
This is trivial. Each change produces new information, so lots of changes produce lots of new information.

2. natural selection must stay neutral during the micro-e process, only to become a stabilizer at the end of the macro-e process.
NO.

So item#1, that macro produces sizeable increases in the genome, representing new information, is based on my belief that there are no intermediate steps for some structures. Take as an example chlorophyll, even though it may not be a good enough example. If there are no light-to-energy conversion molecules lesser than chlorophyll, there are no precursors.
So-called irreducible complexity. The problem with this is twofold: First, no claim of irreducible complexity has ever stood up to examination. Second, biological systems can and do evolve by the simplification of more complex structures, so even if something was irreducibly complex, it could still evolve naturally.

So chlorophyll had to be coded for completely at once.
No. Even if there were no precursors, this still would not be true.

This would be macro evolution unless there was already a similar molecule for some other purpose, that could be changed by a micro-e process. The test for macro-e is that a new sequence of significant size appears that RM would have to work on for too long a period of time in order to get in one fell swoop.
That's not evolution, macro or otherwise. That's magic fairies.

Macro evolution is where 'there's not enough time' because the intermediate steps are too large to be micro-e.
So your "macro evolution" is defined not to be evolution at all then?

Natural selection must remain neutral during the intermediate steps that do not express some advantage.
No.

So in macro-e, it would be best if natural selection would look the other way for a while.
No.

Natural selection can not do anything but stay neutral to help along the intermediate steps.
No.

That is another characterisitic of macro-evolution - being where NS must be neutral while new information is forming.
No.

Your incomprehension of evolution is awe-inspiring.
 
Last edited:

ISF - Join now!

Every member here is approved by hand. No bots, no spam, just people who care about evidence and honest debate.

Membership is free!

Create your free account

Back
Top Bottom