Do the ID interpretations of NFL theorems imply the creationist Genetic Entropy hypothesis?

_{Salvador Cordova

July 5, 2013

Darwinism, specified complexity}

Share: Facebook; Twitter; LinkedIn; Flipboard; Print; Email

The ID interpretation of No Free Lunch theorems argues that Darwinian processes on average will not do better than chance processes for the emergence of biological complexity. As has been debated at UD, it’s not merely a question of what is possible, but what we should reasonably expect. For example, see: The Law of Large Numbers vs. KeithS, Eigenstate, and my other TSZ critics.

The Genetic Entropy hypothesis by creationist John Sanford argues that biological complexity is gradually going out of the human genome and possibly the entire biosphere. I provided cursory analysis that lends credence to both the ID interpretation of No Free Lunch theorems and the Genetic Entropy thesis here: The price of cherry picking for addicted gamblers and believers in Darwinism.

I think if random chance tends to degrade and eliminate biological complexity, and if the No Free Lunch implies Darwinian evolution will do no better than random chance, then the ID interpretation of the No Free Lunch theorem mandates the Genetic Entropy thesis. That is, if complexity on average cannot go up, at best it can be maintained, and will probably go down, hence NFL effectively predicts Genetic Entropy.

I don’t see any way around it. I welcome reader comments if they think this is correct.

[posted by scordova to assist the News desk for 1 week with content and commentary]

Comments

Search spaces whose elements are themselves searches may be thought of as consisting of probability measures or fitness landscapes or other functions on the original search space. Spaces of functions on an original space are always richer than the original space and, in case the original space is finite, grow exponentially or even super-exponentially in cardinality. Spaces of probability measures, for instance, are always infinite.
Footnote 18 from Dembski's, Ewert's, and Marks's article "A General Theory of Information Cost Incurred by Successful Search"DiEb_{July 7, 2013
July
07
Jul
7
07
2013
07:49 AM
7
07
49
AM
PDT}

DiEB:
Unfortunately, the issue which I raised cannot be discussed without at least the most modest understanding of probability theory not only on finite but on infinite sets.
LoL! Your position doesn't deserve a seat at probility discussions and infinite sets do not exist in the real world of biology- infinity only exists in our minds. Not only that but followers of Cantor don't seem to be able to deal with infinite sets. And you don't seem to be capable of ideas. BTW evolutionary algorithms are design mechanisms...Joe_{July 7, 2013
July
07
Jul
7
07
2013
06:56 AM
6
06
56
AM
PDT}

Thanks Dieb, I have nothing much to add, it's not my field of expertise, and hence I posted to solicit more expert discussion. At issue is what we should actually see in the field. I think deterioration is the correct prediction. Salscordova_{July 6, 2013
July
07
Jul
6
06
2013
04:06 PM
4
04
06
PM
PDT}

@Joe, it is very nice of you that you try to contribute. Unfortunately, the issue which I raised cannot be discussed without at least the most modest understanding of probability theory not only on finite but on infinite sets. Our little exchange of ideas let me to the conclusion that you are lacking the basic skills to follow this discussion on an advanced high-school or even undergraduate niveau. So, while I acknowledge your taking notice, it is of little consequence for any of the ideas involved, I'm afraid.DiEb_{July 6, 2013
July
07
Jul
6
06
2013
03:43 PM
3
03
43
PM
PDT}

Anyone what? Anyone notice that DiEB did NOT post any evidence that demonstrates darwinian processes can actually do something more than just eliminate the defincient and defective? That anyone? OK, I have noticed. Satisfied?Joe_{July 6, 2013
July
07
Jul
6
06
2013
02:01 PM
2
02
01
PM
PDT}

Anyone?DiEb_{July 6, 2013
July
07
Jul
6
06
2013
01:55 PM
1
01
55
PM
PDT}

My 2¢: In all these optimization problems, you try to find the extremum of a function f taken from a set of functions F by evaluating f at various points. The NFLT hold when F has a certain kind of symmetry: it has to be closed under permutation. This is trivially given when you look at the characteristic functions of elements of a search space - the favorite example of Robert Marks and William Dembski. But when the functions get less simple, this closedness of F is rarely given: I haven't seen a real world example where the functions take more than two values and F is closed under permutation - other than F being the set of all possible functions. In fact, F seems to be very "asymmetric" in general - think of TSP as an example. Evolutionary algorithms aren't clever or elegant. They tend to be wasteful - and if you have an "intelligent" solution for a problem, this is to be preferred. But they are easily programmable and often work on the sets F we find as problems taken from reality. I don't think that the proponents of ID will find a striking argument for ID by looking at the NFLT - and I'm generally not happy by the way they try to exploit them: see my e-mail to Dembski, Ewert & Marks about BI:NP - A General Theory of Information Cost Incurred by Successful Search .DiEb_{July 5, 2013
July
07
Jul
5
05
2013
02:48 PM
2
02
48
PM
PDT}

Prev 1 2

You must be logged in to post a comment.

Leave a Reply