>>13108871no, you see a pattern, you define what variables/networks are required for a conclusion, then you generate a generic "rule", then you generate exceptions to that rule, if any exceptions (improper logic, unproven variables) are found then its marked invalid.
if no exceptions are found, the rule is theoretically possible and has to be proven right.
rules are combined together to make bigger rules, these can be encoded in to english, math, logic, code, or visual representation or multiple.
rules can be proven rewarding, or "happiness"-causing, but still be wrong.