2003 Jan 18.

John Bracht posted a paper "Inventions, Algorithms, and Biological Design" in which he discusses my ev program.

• Schneider notes, "Remarkably, the cyclic mutation and selection process leads to an organism that makes no mistakes in only 704 generations." But precisely how remarkable is this result? According to Schneider, "The ev model show explicitly how this information gain comes about from mutation and selection, without any other external influence, thereby completely answering the creationists." But has the information gain really been driven only by mutation and selection? On closer inspection it becomes clear that Schneider loaded a tremendous amount of information into his program from the beginning, and that what the program accomplished was not as monumental as he would have us believe.

  The term 'information' in the paper is the information in the binding sites. Careful inspection of the paper and program source code will reveal that this is ZERO at the start of the evolution run. So although the program has information, in the sense that it takes a certain number of bytes of computer memory, that information is irrelevant to the problem at hand. Bracht missed the main point of the paper!

• ... together, these parameters define the class of all possible "organisms" which the program can generate and test, and these parameters were all deliberately chosen by Schneider to produce the complex and interesting result. The fitness function counted as "mistakes" those instances in which a binding site's value is below the threshold or a non-binding site's value is above threshold. Therefore, the fitness function selected directly for organisms that had high binding site values (above threshold) and low non-binding site values (below threshold). However, Schneider's condition of success was precisely the same: to create a set of organisms that had binding site values above threshold and non-binding areas below threshold. He encoded the very conditions of success into his fitness function-and then claimed that his program had created this information from scratch!

  Bracht has again confused initial conditions with the results from the program. The program started out with no information in the binding sites and by pure Darwinian mutation and selection evolved measurable information at the sites. Bracht is invited to try the program. It should give similar results within a wide range of initial parmeters. Of course some parameter settings would disallow evolution (eg no mutation or demand for more information in the binding site than allowed by the matrix). The fact that there are a lot of parameters is a reflection that nature has a lot of variable parameters.

  Bracht might ponder that the program corresponds rather closely what happens in nature.

• John Bracht
  • John Bracht is Managing editor of ISCID. In other words, he is a creationist. (The so-called Intelligent Design movement is old-style creationism in a new fancy cover.)
  • Bracht apparently (according to Gro Mit) started the page Information Theory: Finding the Immaterial (i.e. non-natural) Properties of life The current authors are Casey Luskin and Reid Hankins but presumably Bracht is still responsible for this. Since it is chock full of errors, I'll make a few notes:

    The page quotes Dr. John Baumgardner as saying that "Language involves a symbolic code, a vocabulary, and a set of grammatical rules to relay or record thought. Many of us spend most of our waking hours generating, processing, or disseminating linguistic data. Seldom do we reflect on the fact that language structures are clear manifestations of non-material reality." The fact that is missing in this and the later statements is that information always has a physical carrier. So the supernatural implications they try to create fall flat.

    By the way, the category that is distinct from matter and energy is well known, it is called entropy! So no fancy supernatural explanations are needed. It was a serious mistake and a disservice to the reader for them to neglect to mention this fact. He is also apparently unaware of Shannon's work.

    "The very existence of the code can only be explained through a pre-existing Intelligence who put it there." That this statement if false is demonstrated by many experiments and the ev program.

    "Information does not just "happen" by chance, random processes' That's correct, but they neglect to mention that selction is not a random process. This is a classical creationist tactic and I suggest that you train yourself to recognize it.

Acknowledgments

Schneider Lab
origin: 2003 Jan 18
updated: version = 1.00 of bracht.html 2003 Jan 18