Introduction to models of evolution with respect to skills - (annotated Table of Content)

Working definitions:

From an information-processing perspective, evolution involves copying information:

• There is a target population.
• The population can be described as multiple sets of distinct individuals that are all fairly similar to one another, but not identical.
• The sets can be described as distinct generations, with an odering (typically temporal) from parent generation to child or offspring generation.
• Information is copied from one generation to the next.
• The copying is uni-directional, from parent to offspring.
• The copying is selective, so that there is a higher probability (or frequency) of copying from a member of the population that has higher fitness.
• The selectivity of copying is typically modeled as a function of variation in the population.
• There is variation in the information, and a process of maintaining or increasing the variation.
• The copying may be inexact.
  • The information from two parents may be mixed during the copying process, where the mixing introduces variation.
  • The copying process itself may introduce variations (mutations).

The variation and selective copying leads to changes in the distributions characterizing the variation in the information in the population.

• In the short term
• In the long term, over many generations
• An animated representation of variability, over one generation
• An animated representation of variability and selection, over one generation - for a single optimum for distance-based fitness

An interactive animated simulation of the evolution of a two-dimensional trait over 100 generations: please click on the links in order.

• describing the start-up population: parametric bounds for randomly generating the population
• followed by the attributes of the 'ideal' individual in the population: the optimal attribute combinations - for selection (copying preferences)
• followed by the copy, select, and vary mechanisms: inexact copies with variability, copied independently
• followed by a simulation run
• selecting the stats to display for the generations: too many causes confusion
• Review params: paraneters used in the simulation
• Evolution as information passing with a learning component: experimenting to optimize, using all individuals from a population

An information-processing feasibility model is successful if it can simulate the phenomena to be described, within the simplifying assumptions that have been made explicit, and without violating major constraints that should also have been made explicit (such as not requiring a racial consciousness, and not requiring extra-sensory perception).