19 March, 2009

Of Weasels and Wankers

I'm currently reading Richard Dawkins's The Blind Watchmaker (which is a book guaranteed to make creationists sob). By computer standards, the book is ancient - the computer he wrote it on had a memory measured in kilobytes. Nowadays, of course, that's laughable.

But he still managed to get that wee little machine to sit up and do tricks. One of his tricks was the Weasel program. It's a simple, elegant display of random mutation vs. random mutation with selection. It all had to do with monkeys writing Shakespeare, and the power of evolution to bring order from chaos. In just a few dozen generations, selection causes random mutations of letter strings to converge on "Methinks it is a weasel." The need for infinite monkeys, time and typewriters is eliminated when you throw a little selection into the mix. It's an outstanding tool for understanding how a little thing like natural selection combined with a bit o' mutation can produce all the brilliant complexity life demonstrates.

This, apparently, has made creationists sob for over 20 years. In fact, William Dembski, hereafter more aptly named Dimski, just can't let it go:
Over at uncommon descent William Dembski is musing over Richard Dawkins Weasel program. Why you may ask?

[snip]

Such is its power, the Issac Newton of Information Theory, William Dembski, spent a not inconsiderable portion of his time attacking this toy program. In particular, he claimed that after every successful mutation, the successful mutation was locked into place, and couldn’t be reversed. But he was wrong, and it seems he just can’t admit it.

As you can see, by using the Courier font, one can read up from the target sequence METHINKS*IT*IS*LIKE*A*WEASEL, as it were column by column, over each letter of the target sequence. From this it’s clear that once the right letter in the target sequence is latched on to, it locks on and never changes. In other words, in these examples of Dawkins’ WEASEL program as given in his book THE BLIND WATCHMAKER, it never happens (as far as we can tell) that some intermediate sequences achieves the corresponding letter in the target sequence, then loses it, and in the end regains it.

Selection strikes again! In the book, instead of plumping for 160 pages of gibberish, Dawkins presents us with every 10th fittest string. Dawkins even kindly provided the rules he gave the program. Someone who isn't a complete Dimski can whip up a version for themselves, and watch the strings wander all over the place, even indeed losing and regaining target letters, before landing on the target sequence. Indeed, there's even a BBC documentary in which Dawkins puts his program through its paces, and you can watch it happen.

A normal person would go, "Whoops." Dimski screams "Conspiracy!"
That leads one to wonder whether the WEASEL program, as Dawkins had programmed and described it in his book, is the same as in the BBC Horizons documentary.
What a doofus.

So what does a Dimski do? Puts a "chief programmer" on the case:

In any case, our chief programmer at the Evolutionary Informatics Lab (www.evoinfo.org) is expanding our WEASEL WARE software to model both these possibilities. Stay tuned.

Say what? How long does it take someone to write a program that basically takes a string, copies it with mutations, compares it to a target, chooses the best mutant, copies and mutates the new string, and compares again until the target is reached?
Well, this is Dimski we're talking about.

*cue Musak*

Doop-de-doo. Dum dum dum. Hmm, I wonder what's mutating in the fridge?

My goodness, how that grass does grow.

Whelp. Let's check in and see how things are going:
It is now 62 hours since William Dembski posted that the Evolutionary Informatics lab was going to try and reproduce Dawkins Weasel Program according to how it was actually written, as opposed to their fantasy version. In that time I’ve resurrected an elderly program, and several readers have made their own weasels from scratch. Commenter Anders has even made a Python version that puts “freely mutating” and “locking versions” head to head with great graphs.
Y'know, Billy Dim, I do believe it's taking you so long because the program you're trying to create never existed. Makes it rather hard to duplicate, dunnit? It's kinda like trying to selectively breed for gryphons by way of proving all the myths are true.

This little dustup does demonstrate a rather powerful truth, although not the one Dimski intended. It's really hard to produce scientific results backing up claims that are just flat-out wrong.

Which goes a long way towards explaining why IDiots haven't managed any scientific proof yet, eh?

1 comment:

Cujo359 said...

I suspect a competent Python programmer could have knocked either version of those programs out in an afternoon. I think Dr. Dimski needs to hire some new programmers.