CF011: Design in evolutionary algorithms

Claim CF011:

Source:

Response:

1. The fitness function of genetic algorithms need not include any new information. A fitness function can be expressed as whether the algorithm performs better or worse in a particular environment. The only information is provided by the environment, which is usually modeled on the real world. The claim makes sense only if design is defined as what is in nature already.
   
   One may argue that nature and design are inseparable (and Dembski seems to make just such an argument; Dembski 2002, xiv), but this invalidates the design argument. Design only has meaning if contrasted with nondesign, and defining design as all of nature makes nondesign nonexistent.
   
   
2. Genetic algorithms often come up with novel solutions which sometimes even surpass direct human designs (Koza et al. 2003) and which do not rely on human expertise (Chellapilla and Fogel 2001). Humans may have told the algorithms what to do, but it is the how that defines the design.
   
   
3. Genetic algorithms are not perfect evolutionary simulations in that they have a predefined goal which is used to compute fitness. They demonstrate the power of random variation, recombination, and selection to produce novel solutions to problems, but they are not a full simulation of evolution (and are not intended to be). In simulations of biological evolution, fitness is evaluated only locally; survival and reproduction is based only on information about local conditions, not on ultimate goals. However, the simulations demonstrate that distant fitness peaks will be reached if there are conditions of intermediate fitness (Lenski et al. 2003). Evolutionary processes do not "search." They respond to local fitness topography only. The fact that evolution (occasionally) reaches fitness peaks is a by-product of evolving on correlated fitness landscapes using purely local fitness evaluation, not an intended outcome.

Links:

References: