I guess I should clarify that I suspect some complex systems- like human biology and advanced AIs- are effectively unknowable even to the smartest people. We can craft a narrative this is sort of a projection of the true complexity down into a linear space that makes rational cause-and-effect sense to us ("lock and key hypothesis", counting up enthalpic contributions of a ligand/protein pair), but I suspect much of our failure in drug discovery comes from missing complex higher order interactions, subtle feedback and regulation effects, and unknown population or environmental variability across human biology, and those really are out of the reach.
Although it is still very controversial, I think we should start thinking about whether it is sensible (both technically and morally/ethically) to think about human brain augmentation (either by working closely with an AI that helps us understand the complexity, or by actually improving how our brains work on the physiochemical level) to improve our understanding of human biology (sort of a recursive self-improvement loop of improving the brain to learn how to improve the brain more). It's still very scifi, and society has previously shown extreme resistance to this sort of self-improvement, but I think we are on the cusp of a societal phase transition where the most adventurous of us will be rewarded for taking risks.