Evolutionary Algorithms and ID

Let me quote a small paragraph taken from the monograph « introduction to evolutionary informatics » written by Marks II, Dembski and Ewert. Here it is:
« AVIDA, EV, Dawkin’s WEASEL problem, and meta biology are all written by proponents trying to demonstrate Darwinian evolution works. Each is a designed stochastic process that, like Buffon’s needle or tetherball, converges to one or more fixed points. To do so, each requires sources of knowledge to generate active information to guide the search. The success of the program depends on an intelligent designer. Conservation of information requires it ».
That’s something to think about, isn’t it?

Are you telling me you read a few thousand papers and all the papers they cite in a span of a few hours to come to that conclusion? Ok buddy


It is remarkable that this little snippet has salient factual errors. Pretty surprising actually!


No it’s ridiculous to be honest.

Think about it just a little bit. Someone comes along and says evolution doesn’t work without being front-loaded some sort of goal. So you sit down and write a piece of software in which evolution happens, but there are no goals anything evolves towards. Aha, your interlocutor says, but the simulation itself was written with the goal of showing evolution can work without a goal, so it has a goal anyway.

That’s dumb. Just plain dumb.

What evolves, evolves without a goal. The simulation is not set up to evolve towards anything in particular. The person who wrote the software intended for the software to achieve something, yes: Demonstrating that the evolutionary process can occur without having a goal set up for it(it is not being guided towards some particular organism or adaptation). Or he might even have intended for it to accurately reflect real evolution, and for it to become a research platform that might give insights into the phenomenon of life, or evolution.

What you’re saying is basically that if I claim the weather has no goal, and I write a piece of software that simulates the weather as a contingent process without a goal, then because I as a coder had the goal of showing that the weather can work without a goal, then the weather being simulated has a goal too anyway. The goal of being goalless.

This is the last refuge of someone who lost an argument. Come the fork on.


The paragraph below is taken from the website « evolutionary informative lab ». I though it may interest you in the context of your ongoing conversation on MN.


Digital Evolution Intelligent design is the study of patterns in nature best explained as the product of intelligence. So defined, intelligent design seems unproblematic. Archeology, forensics, and the search for extraterrestrial intelligence (SETI) all fall under this definition. In each of these cases, however, the intelligences in question could be the result of an evolutionary process. But what if patterns best explained as the product of intelligence exist in biological systems? In that case, the intelligence in question would be an unevolved intelligence. For most persons, such an intelligence has religious connotations, suggesting that it as well as its activities cannot properly belong to science. Simply put, intelligent design, when applied to biology, seems to invoke ‘spooky’ forms of causation that have no place in science. Evolutionary informatics eliminates this difficulty associated with intelligent design. By looking to information theory, a well-established branch of the engineering and mathematical sciences, evolutionary informatics shows that patterns we ordinarily ascribe to intelligence, when arising from an evolutionary process, must be referred to sources of information external to that process. Such sources of information may then themselves be the result of other, deeper evolutionary processes. But what enables these evolutionary processes in turn to produce such sources of information? Evolutionary informatics demonstrates a regress of information sources. At no place along the way need there be a violation of ordinary physical causality. And yet, the regress implies a fundamental incompleteness in physical causality’s ability to produce the required information. Evolutionary informatics, while falling squarely within the information sciences, thus points to the need for an ultimate information source qua intelligent designer.

I thought that you were trying to turn the conversation back to your claim:

@gpuccio has precisely shown that such complex proteins with very high FI (>500 bits) exist. Now the question is whether their evolution by the traditional RV+ NS mechanism is plausible. The answer is no.

Please show that the answer is no for immunoglobulin VDJ recombination.

I’d just like to add that Gpuccio has not shown that any known protein has 500 bits of FI or more, because the method he uses by which to gauge the number of sequences above the minimal threshold for function doesn’t accomplish what he wants it to.

I do have to agree with @Giltil that I don’t understand why the thread Gpuccio posts in was locked.

@Giltil, do you see a difference between a creature’s design (humans, animals, aliens) and the Creators’ design? Are these the same category or not? Do you see any distinction here at all? Or do you think God is in the same category as humans, animals, and intelligent aliens?

1 Like

AKA The Third Law of Creationism:

1 Like

Ahh yes, computers and the programs that run on them are designed, so everything that happens in the program was designed too. Of course. The third law of creationism can basically be restated as: Nothing that happens in a computer can ever amount to evidence against design.


Does water running downhill have a goal? Or is it simply the entropy of a system progressing to a maximum? I have used an EA for a constrained optimization problem, which is the same sort of maximization.

1 Like

I guess they are differences between the creator and human designs. But no doubt they are also similarities, among which figure purpose, knowledge, intelligence, foresight.

Sure there are similarities. I am just convinced that the differences are far to large to neglect. We have to have a good handle on the differences, and expect that our intuitions about creaturely design will not always carry over to divine design.

Don’t you agree?


I pretty much agree « that our intuitions about creaturely design will not always carry over to divine design ».
I would risk to say that, unlike human design, God design has something to do with the performative power of his word (God says something, and that thing comes into being).


I don’t see much point in calling that “design.” I’d say that what I see in the workings of biology is nothing like human design.


What about this
image https://thumbs-prod.si-cdn.com/tMNPTIk69FleN-rjezG6WYWKh1E=/420x240/https://public-media.si-cdn.com/filer/Surprising-Science-nature-gears.jpg

What about it? What does it do? Periodicity is very common in nature.

Did God design this periodic structure?

How is your calculation of the FI for VDJ recombination going? Can I help with something?


Periodic phenomena in nature are not contingent and therefore cannot be attributed to design.

Okay, let’s try.
So, in order to calculate the FI for VDJ recombination, the first thing to do is to define the function. I think we can say that the function of VDJ rec can be stated as follow: « produce as many antibodies as possible ».
Once you have defined the function, the second thing to do is to calculate the probability that the system will find the target, or, put differently, will produce the function. So what is the probability that the system « will produce as many antibodies as possible »? The thing to note here is that the system is precisely configured/organized to produce the function. Therefore, the probability that VDJ will produce the function is very high, nearly 1. According to this analysis, nearly no FI is produced by VDJ recombination per se and the high FI associated with antibodies preexists it.
I realize that this result may seem strange to you. In that case, you could ask yourself the following question: what is the probability that a car factory will produce cars?

Not really. I’ve read Marks et al’s work on GAs, and their papers on Tierra and Dave Thomas’s Steiner GA could and probably did get written without them having run those programs even once. They don’t quote any results at all from any of their own GA runs, they just analyse (badly) the results of other people’s runs.

They also make some basic errors. For example:

  • WEASEL wasn’t written to demonstrate evolution, only to demonstrate selection.
  • Avida doesn’t converge to fixed points
  • Buffon’s needle is a mathematical problem; it doesn’t converge on anything (though simulations developed to solve it can).
  • Their list (deliberately?) omits Tierra, which unlike the others has no “source of knowledge”
  • They can’t even get Dawkins’ name right!

You’d learn more about GAs from experimenting with Tierra or Avida or others for a couple of hours than you ever would from reading Marks’ and Ewert’s opinions.

P.S. Dave Thomas comments on some of their work here, exposing their incompetence/mendacity:
“the problem my algorithm was solving, Steiner networks, was “replaced” with a much simpler problem, Minimum Spanning Trees. This ruse enabled Ewert to launch a (straw) attack on my genetic algorithm for solving Steiner’s problem.”

1 Like

Wrong. Trivially, laughably wrong. You haven’t included anything about the antibodies being different or having binding sites, without which the antibodies would be of no benefit…