#10199: — 12/03  at  07:54 AM
I would like to see these people first demonstrate their understanding of evolution by simply showing that they are able to apply their knowledge by creating a genetic algorithm that then solves some search problem for optimal solutions.

Then they could point out how evolution works there and how it all is actually a careful plan by some agent and nothing else.

They can't say genetic algorithms do not exist. One can be shown to them.

It is a simple enough a request to show that if someone knows how something works, they can make it also.

They should therefore be able to explain how it works.

How are they going to plug an agent to a deterministic system there?

#10202: — 12/03  at  09:39 AM
Vince--genetic algorithms are created by intelligent designers (computer scientists). The parameters of the algorithm are set by the designer to ensure that the final "population" has good "fitness". It's really not like evolution at all; it's just stochastic search with evolutionary terms thrown in because they are somewhat intuitive, but mostly because they were fashionable at the time it was invented.

#10220: Bryson Brown — 12/03  at  11:37 AM
Ben--the formal parallels between genetic algorithms and evolution in organisms are pretty straightforward. Dembski had to get very slippery in No Free Lunch to avoid these arguments (he assumes that fitness is arbitrarily distributed across the search space-- a ridiculous assumption for real organisms in real environments). Any formal model of evolution will be the product of human modellers. But it can still prove that variation-and-selection can arrive at complex, and even 'irreducibly complex' systems.