Side Comments on Algorithmic Specified Complexity

It’s been a long time since I read Mark’s paper, but IIRC ASC is an approximation to Kolmogorov Information - is that correct?

Nope. It is somewhat idiosyncratic, and rather than get into a difficult word parsing exchange, it will be easiest for @EricMH and I to come to an agreement upon implementation that we can analyze.

Thanks - I’ll re-read the paper … somewhere in my endless free time …

1 Like

Glad you agree that ASC is not innovative. And that ID is not innovative either. Realize that ID theory is not settled mathematically and nothing in ID has been proven elsewhere by more eminently qualified people.

Dr. Swamidass is correct. And more importantly Claude Shannon was correct in 1948 when formulating the information theory of communications which defined entropy and mutual information. You are correct that our entire modern communication infrastructure relies on Shannon’s insights regarding mutual information.

No ID, ASC and CSI have not been linked to natural processes in any way. Shannon’s Information Theory remains intact and the Theory of Evolution remains intact. ID, ASC and CSI has added nothing of value to either.

No information theory tells us nothing about whether God exists and whether humans have immortal, immaterial souls.

But other than that, it was nice to study information theory again. And I am glad to have met Claude Shannon and to hear a lecture from him along time ago. Have a nice day.

1 Like

Don’t see it. Lots of hand waving.


@patrick please stay out of this thread. There is a side comment thread for things like this.

Each step directly follows. You are being hand wavy about the hand waviness. If you are interested in understanding, please state which steps are “hand wavy” and I can clarify.

I will defer to Dr. Swamidass to take apart your arguments.

Yes, it seems that way to me.

In particular:

I very much doubt that assumption.


Good point. Human intelligence is natural, and it is neither deterministic nor random.

@EricMH help us out?


This is such a important point I’m tempted to open a new thread. All the proofs for information also apply to the human brain, and also human intelligence. We are made up of natural processes. If we agree that humans can produce the “information” that ID speaks of, why exactly can’t other natural processes? Is embedded I need the argument that human intelligence requires an immaterial soul? Is that true for animal intelligence too?


First problem is the word “natural”. Without a precise definition we could say God is “natural,” and solve all our disagreements by redefining our terms. Naturalism becomes coexstensive with theism, so the term is a distinction without a difference.

I define a “natural process” to be everything that operates by the laws of physics. The laws of physics are circumscribed by chance and determinism, which in turn can be circumscribed by a Turing machine with a random oracle.

This, by my argument, implies that anything that can create mutual information cannot be reduced to physical laws. So, if human intelligence can create mutual information, which it seems to do with great regularity, then it cannot be reduced to physical laws either. Thus, the human mind must transcend the physical world.

This is probably true for animal intelligence, too, since they exhibit purposefulness. In fact, all of life, including plants, seems to be purposeful, so no form of life can be reduced to physical laws. This implies some form of vitalism.

With that definition, my conclusion would be that there are no natural processes.

Well, okay, that’s not quite right. We do construct abstract models that operate by the laws of physics. So, based on your definition, I would have to conclude that our abstract models are natural processes, but nature itself is not.

This seems backward.

I do understand that you probably disagree with me. Quite a few people do. Perhaps this part of the discussion should move to Science and Philosophy, since it has to do with the nature of physical laws.

Our engineers build heat seeking missiles. They are built based on the laws of physics, and they use computers to control them. The pilot of an airplane being targeted by one of these missiles would surely think that they are behaving in a way that seems purposeful.

In my opinion, these heat seeking missiles are, indeed, purposeful. And their purposefulness was built into them using physics and engineering. No vitalism was required.


Easy. Natural is all that God created.

That sounds like an assumption. Even in humans have immaterial souls, it sure looks that human intelligence emmeninates from the natural mechanisms of material brain. If you think that human intellinence can produce “information” but natural process can’t, sounds like a contradiction.

What do you mean by reduced?

Emergent properties cannot be reduced to physical laws but they are still entirely explained by the complex interactions of physical laws. They are a consequence of physical laws but not fully explained by them.

None of this implies design.

What about viruses? What is their teleology?

Here you are going far afield. For me.

What is your personal views on origins? Do you affirm an old earth? Do you affirm common descent? Are you YEC? No judgement.

4 posts were split to a new topic: Do Heat Seeking Missiles Have Teleology?

Are angels natural? Are immaterial human souls natural? If so, our disagreement is minimal.

However, that is not the normal naturalist argument. The normal argument is that the natural world is all that exists. God, angels and souls are not natural, thus do not exist. Clearly, this argument uses a different definition of “natural” than you do. What is this definition of “natural” used by naturalism?

Naturalists seem to define nature as everything that reduces to the laws of physics, and then claim everything is natural, which is just incoherent in so many ways.

The human brain could be an “antenna” that allows our immaterial mind to interact with the physical world. Assuming our mind reduces to the physical operation of the brain because impacting the brain impacts the mind is fallacious. For example, watching a sad movie makes my mind sad, thus sad movies impact my mind. However, I do not then conclude my mind is the movie I am watching.

You seem to contradict yourself here. In my mind “reduce” is exactly the same as “entirely explained by the complex interactions of physical laws.” Perhaps you can clarify.

I don’t know much about viruses, but they sound like machines from the little I know, not living entities. This in turn makes sense of their tendency towards disorder, since teleological entities seem to move towards order.

I don’t understand the relevance of these questions for the discussion at hand. Any answer I give appears orthogonal to the question of whether humans are reducible to matter. God could have specially created a bunch of robot humans in 6 days, or guided the development of soul vessels through bililons of years of evolution.

You have been here for a while and we care what people think. Personal beliefs aren’t relevant in a scientific argument, but they are in friendship. Let us know what you think.

1 Like

Since the main thread is closed I will add my comments here.

Continuing the discussion from Eric Holloway: Algorithmic Specified Complexity:

@swamidass: You may understand this better than I do, but for the benefit of others:

In practice there will always be some optimal way to compress for a particular instance of X. In theory we are dealing with a random distribution of X, and no single compression scheme can always be optimal.

Technically we should also count the length of the compression/decompression code as part of the length, but when we consider strings of of arbitrary length that becomes a very small fraction which can be theoretically ignored. I like the analogy of a self-extracting ZIP fille, which contains the information needed to decode itself. In theory the compressed file should have reduced length, but is practice this may not always be true.

One more: In theory we can have fractional bits of information, but in practice we must always round up.

1 Like

From above …

  1. We can mathematically prove mutual information cannot be generated by determinism + randomness.

Are you sure about that? Because in statistics mutual information is covariance. If we start with a deterministic relation Y=X and add a random error term \epsilon so that Y=X+\epsilon, then covariance describes the shared information between X and Y. How is the not mutual information generated by determine plus randomness?

  1. Therefore, natural processes cannot produce mutual information.

Natural processes cannot generate correlations? No, that’s wrong, and counter-examples are everywhere.

It is a short step from 5 to get to the immaterial soul and the existence of God.

Every time I see one of these proofs, I have to wonder about the theological consequences of reducing God to an equation.


Someone please confirm I am understanding the directionality of ASC correctly. Given …

ASC(X, P, C) = -\log_2(P(X)) - K(X|C)

If X is non-compressible, then K(X|C)=0, and ASC is maximized. This also means X is completely random. ASC seems to be detecting randomness, not design or the work of intelligenge. Is that correct?

I note this is opposite from how Dembski defined CSI in his 2005 paper, where a fixed string (non-random) would maximize CSI.