Kevin Dorner

<a href="http://www.infidels.org/~meta/getalife/" target="_blank">Get-A-Life</a>... not a request but instead the article's title.

Dr Rick

The fallacious nature of the challenge posed in the opening post leading to the flawed conclusion that evolution is merely "a guess" can be illustrated by re-wording the former:

Name one technological advance that is the result of [Relativity].

Rick

[ August 16, 2002: Message edited by: rbochnermd ]</p>

lpetrich

Vanderzyden, you have either neglected to read or chosen to ignore what Baloo had posted on the first page of this thread:

Genetic algorithms.

These are inspired by evolution by natural selection.

Genetic algorithms are a function-optimization algorithm, one of a large number of such algorithms. I am familiar with several of them, having worked on code that implements them. Many of these algorithms are for special cases of functions, like continuous functions, differentiable functions, and so forth, where the special-case features make possible algorithms adapted to them. How might we do this in the most general case?

A simple way is to take some initial parameter set, make a random change to it, and accept it if it improves the function value. This is natural selection at its simplest, and it does work, though it can easily get stuck in local minima.

One can get around being stuck by accepting a changed parameter set with a certain probability if it makes the function value worse; that enables the optimizer to back out of weak local minima.

And one can improve one's confidence in one's result by optimizing several parameter sets; and then selecting whichever ones gave the best results.

This method can be improved by deleting the worst-performing parameter sets and making extra copies of the best-performing ones, so as to better probe the parameter space near the best ones' values.

And improved even further by doing crossover between parameter sets, so that there will be a certain likelihood of two improved pieces of parameter set being brought together.

Vanderzyden, I suggest that you look at the genetic-algorithm literature some time.

Except that such simulations have actually been done.

Why is that?

And your point is...

Vibr8gKiwi

I see people have already responded to your request for more information on genetic algorithms and evolutionary algorithms, but here's my comments.

These algorithms do relate in more than the "slightest" way to biological evolution as that was what they were modeled after. The technique even defines software "genes," the resulting phenotypes the genes produce, as well as operations on the genes to simulate duplication/reproduction, mutations, crossovers, etc. There is also a higher level where an environment is simulated where populations of the genes are reproduced and non-randomly selected (with a selection function ala natural selection). The result is an engine well suited for solving solutions to particular types of open-ended problems.

It's interesting to note that what boils down to a simulation of biological evolution is used on computers to generate solutions that appear "designed" yet are simply the result of a process of directed (yet mindless) trial and error. Especially since I suspect this is exactly what you think evolution cannot do.

If you're interested in further information:
<a href="http://www-2.cs.cmu.edu/Groups/AI/html/faqs/ai/genetic/top.html" target="_blank">Genetic Algorithms FAQ</a>

[ August 16, 2002: Message edited by: Vibr8gKiwi ]</p>

Albion

And where do you come by that sweeping statement? You're really so familiar with the research on evolutionary theory that you can be sure that it doesn't have a sound mathematical basis? You might want to get hold of a textbook on population genetics.

Or maybe check out some of these abstracts:

<a href="http://www.ncbi.nlm.nih.gov:80/entrez/query.fcgi?db=PubMed&cmd=Display&dopt=pubmed_pubme d&from_uid=11238990" target="_blank">http://www.ncbi.nlm.nih.gov:80/entrez/query.fcgi?db=PubMed&cmd=Display&dopt=pubmed_pubme d&from_uid=11238990</a>

<a href="http://www.ncbi.nlm.nih.gov:80/entrez/query.fcgi?db=PubMed&cmd=Display&dopt=pubmed_pubme d&from_uid=11969389" target="_blank">http://www.ncbi.nlm.nih.gov:80/entrez/query.fcgi?db=PubMed&cmd=Display&dopt=pubmed_pubme d&from_uid=11969389</a>

<a href="http://www.ncbi.nlm.nih.gov:80/entrez/query.fcgi?db=PubMed&cmd=Display&dopt=pubmed_pubme d&from_uid=9679319" target="_blank">http://www.ncbi.nlm.nih.gov:80/entrez/query.fcgi?db=PubMed&cmd=Display&dopt=pubmed_pubme d&from_uid=9679319</a>

Mageth

Software does not exist on the computer, but rather in the minds of software engineers.

I'm a software engineer. I assure you all the software I've ever worked on exists solely on the computers on which it resides, and not in my mind.

Vibr8gKiwi

This statement is so odd as to be hilarious, and its made even funnier by the ironic "think about it" intro.

I assure you that if I was to die tomorrow and my mind was to be destroyed, all the software I've written would continue to exist without me. Also the algorithms will continue to work whether you believe they should or not

[ August 16, 2002: Message edited by: Vibr8gKiwi ]</p>

Mageth

Not only that, but software a programmer writes may produce results (output) which the programmer's mind couldn't produce (e.g. due to complexity of computation), or even predict. Not many programmers have a "bug" in mind when they write a program, but virtually all programs have bugs, many of which can produce quirky, unpredictable results.

[ August 16, 2002: Message edited by: Mageth ]</p>

Starboy

Vanderzyden,

I am a software engineer. Trust me, my clients expect the software I write for them to continue to exist and work on their computers after I am gone.

Starboy

Baloo

I thought I had made things very plain and simple. Let me try to dumb it down a bit.

Vanderzyden asked a very specific question: What benefit to mankind has been the result of evolutionary inquiry?

Answer: Inquiry into the theory of evolution has led directly to attempts to simulate the process with software.

Attempts to simulate the process led directly to the rapid development of genetic algorithms.

The incorporation of these algorithms into operating system design has led directly to much more efficient load-balancing on the part of operating systems on desktop PC's.

This has led directly to faster desktop PC's.

You, Van, are using a PC that is faster than it would be without genetic algorithms. You, Van, a member of mankind, have been directly, demonstratably, and irrefutably benefitted by evolutionary inquiry.

I feel I've led the horse to water... Van, would you like to try providing us with the answer to your own question?

What benefit to mankind has been the result of evolutionary inquiry?