Abductive Reasoning and Problem Solving
michael cook, Analyst, United States, Member
You could try this:
One basic paradigm in Artificial Intelligence (AI) is "Generate and Test", i.e., given a problem, find a solution by generating every possible solution then testing to see if it is correct. For all but "toy" examples this method is impractical because of combinatorial explosion.
A refinement of this method is: "Plan - Generate - Test", in which the search space for possible solutions is intelligently pre-selected.
describes discussions about this sort of thing in the context of a molecular biology laboratory discussion run by Nobel Laureate Joshua Lederberg, who also had a long standing interest in artificial intelligence and made contributions to that field as well.