User talk:DiEb

Slightly late welcome.--Bob M 19:19, 21 August 2009 (UTC)

Thank you very much! DiEb 21:34, 21 August 2009 (UTC)

Dawkins's weasel
A slightly techier version: Totnesmartin 22:19, 21 August 2009 (UTC)

I intend to write a little piece on Dawkins weasel - including some pics. It would be ideal material for RW, I suppose, but for the meantime, I don't know exactly where to put it. I even thought of aSK, but I'm not allowed to up-load pics over there. DiEb 13:29, 22 August 2009 (UTC)
 * You could put it in your ursespace here, which would have the advantage of keeping it free from edits by others. Our copyright is the same as RW's. You'd need to remove it afterwards though unless there was an obvious English-teaching angle. (I seem to recall that there is one theory about the origin of language which is that language evolved in order to allow us to deceive each other more efficiently.)
 * Alternatively I suppose the best place would be liberapedia where any editing would probably be on-message and which also seems to have the same copyright.--Bob M 13:57, 22 August 2009 (UTC)
 * Thanks you for the information - and the nice roof over my head :-) The obvious English-teaching angle would be at first that it is written in the very bad English of a not-native speaker DiEb 14:48, 22 August 2009 (UTC)
 * The thing that struck me (and of course others) from the start was the goal orientation of the program. It leads people to assume that Dawkins thinks that evolution has a goal. Toast 14:59, 22 August 2009 (UTC)
 * It was inspired by the Hamlet-writing monkeys - and shows how evolutionary techniques can make their task easier. BTW: the monkeys have all the fun, but I pity the lectors who have to check their works, only to find out that there is a mistake in one of the last words... DiEb 15:10, 22 August 2009 (UTC)

I just tried to get the following comment through at Uncommon Descent:

Just to summarize: i. Dawkin's algorithm seems to have a low mutation rate between generations, while Dembski's rate is much higher: Compare the first two generations of the first example by Dawkins

1) WDLTMNLT*DTJBKWIRZREZLMQCO*P

2) WDLTMNLT*DTJBKSIRZREZLMQCO*P

with Dembski's example:

1) SCITAMROFN*IYRANOITULOVE*SAM

2) OOT*DENGISEDESEHT*ERA*NETSIL

(changes in bold)

ii. Dawkin says that his algorithm doesn't use latching, while Dembski's algorithm latches.

iii. Dawkins's describes the use of a population (though of unknown size), while Dembski uses only one string

iv. Dawkins gives 43, 64 and 41 as the number of trials his algorithms uses in three runs. The expected number of trials for Dembski's algorithm is 104.55

Methinks it is another weasel.