I think the first thing we should do is to understand how superlative human intellectual performance as it exists today even works, and try to make it commodity knowledge taught in everyday classrooms.

In a typical physics, mathematics or engineering curriculum, we learn a lot of fields, such as mechanics, electrodynamics, calculus, signal processing, control theory, and so on. But we don't learn about the human cognitive processes that went into the discovery of those fields. At best we get a caricatured view of the subject's history. Thus, you might learn that Galois came up with the idea of groups. But what was he thinking when he did that? What were the challenges that he was up against that inspired his idea? In short, how did he think, and how might we emulate him for our problems? Ditto for Maxwell, Riemann, etc.

In all the courses that I took, there was none that explicitly taught me how the mind solves challenging problems or comes up with creative solutions. This knowledge is supposed to be somehow imbibed indirectly by learning the fields themselves, which I think is very inefficient. We need to make dissemination of this wisdom more systematic.

A parallel is the field of strength training, in which specific exercises, equipment, and nutritional methods were discovered over a brief period of time. The professional body builders of today are giants compared to their peers over a century ago. Something similar needs to happen for cognitive skills as well.