Before writing and numbers, skilled behaviour was described in stories or acted out in plays. After the advent of numbers and mathematics we have used numbers to make comparisons easier. (Note: we can see the difficulty of comparisons with pay-equity based on verbal job descriptions.) Since computers and artificial intelligence, we have found more complete and more satisfactory representations of skills through simulations.
We have three approaches to simulating skills in humans, each with its own stregths and weaknesses. The first focuses on a demonstrating the feasibility of simulating the use of the skill in many situations and contexts. The second focuses on simulating the development of the skill, i.e. simulating the maturation of an individual. The third focuses on the evolution of the skill across many species, and across many generations of the same species.
Software engineering is about developing computer algorithms to solve some problems for humans. Some of these are fully automatic, but most have to interact with individuals. For the interaction with some programs, individuals can be found that have all the skills required for the man-machine interaction. In most cases the individuals have to be trained to operate the computer, i.e. skills have to be developed for these operators, and then the operators have to learn the skills. The problems that can be addressed with man-computer collaborations are limited both by what the computer can be programmed to do and by what skills the operator can learn. Skill engineering is about managing and designing the man-computer interface with skills that humans can perform satisfactorily. Learning the skill may also involve a man-machine collaboration.
If we can simulate the human operator as he or she operates the computer, then we have a model of the skill. In other words, we have an AI computer operate the second computer to simulate the man-machine collaboration. The same kind of simulation can be done, where the AI computer learns how to operate he second computer.
With those two simulations we can go through a great variety of operating scenarios to make sure the collaboration works. Three mile island was a good illustration of the collaboration failing under an unusual operating scenario.
We start with the assumption that we have a working simulation, and go through the questions we are trying to address with the simulation.
Innate skills are static. Other skills such as chess are learned, but are static at a moment in time, such as for a game. Let us assume we have a working simulation of a skill and a stuation where we are working with individuals on the same skill. For both innate and learnt skills there are a number of questions that can be asked of the model to help address skill engineering challenges in working with the individuals to get optimum performance using the skill.
Skill-learning models can help us improve the learning process by optimizing the sequence of learning situations and through early recognition of error in learning.
Skill-evolution models might help us shape the direction a target skill is evolving and help us recognize and prevent some possible misdirections. Anticipating the direction skills evolve helps us educate individuals to be prepared for use of their skills 20 to 40 years in the future, rather than becoming outdated.