
Originally Posted by
MensSuperMateriam
DA does not work in this way. It does not capture all the information in a single taken snapshot (or multiple-taken, but one in each step) of HP. DA perceives patterns (changing information) more than the information itself, which are commonly associated to algorithms (the way this information could evolve) and therefore the "IF-THEN-ELSE" alias DarkAngelFireWolf69 suggested.
If you capture a flow of information, you do not know (or implicitly have, although still undiscovered) which the value of the information (the solution of the problem) is. You need to fill that "empty pattern" with concrete information, run the algorithm with concrete values for the variables it uses.
The users of DA, being naturally aware of this dependence between the output (conclusion) with the input (initial data) will understand every solution as no much more than a particular solution, which is not by default "more correct", so to speak. Hence the trend to relativism, disliking for general conclusions, etc.
Being patterns empty of information, they can easily be combined in multiple ways, like pipelines, gererating multiple solutions for multiple (or even single) conditions. At least, the "correctness of a solutions" is seen in probabilistic terms. Using again the pipeline analogy, which carries a higher flow of water would be the more correct.