Artificial ideation?

Logic and science have an impressive track record. And so do intuition and art. But whereas logic is used in, say computer hard- and software, where "correctness" is key to operation, intuition is used in design to generate new ideas, where there is no necessity to be logical, provable or correct. And because logic-based models require theory, they work for big-data analysis, long-term predictions, or puzzle problems (where there's only one "correct" way to solve the problem). But ideation and innovation do not need theory - they are event-based processes and work for short term predictions (often with incomplete information), or for fuzzy problems (where problem solving allows for ambiguity, skepticism and/or corrective action). And as many design problems are fuzzy in nature, involving human diversity, complexity and unpredictability, it seems intuition will continue to surpass logic for "wicked" problems in the design domain. ["wicked" problems have been contrasted with relatively "tame", soluble problems in mathematics, chess or puzzle solving, see Rittel & Webber 1973].