i read somewhere about kasparov vs deep blue, paraphrasing "kasparov didn't win by exploring every moves (he won't have time), he won by ignoring the ones which don't work"

so maybe if ideas are treated as "spam", one needs to develop good filter around it