Quantity is Quality

In 1981 when I found out how cheaply microchips could be mass produced, it seemed EVERYTHING should be made from microchips. Walls, floors, furniture, sidewalks, roads . . . all passive objects should be THINKING or at least performing useful calculations.
Once enough stuff is thinking, it will inevitably start solving problems (its own or ours).
This is known as an intelligence explosion. With enough computational capacity, any solution can be found automatically. The best solution to stupidity is brute force, both physical and computational.

This process may yet save us all. Perhaps nothing else could save us, at least not in a real world filled with empowered stupidity.
L. Ron Hubbard certainly didn't rely on his followers' judgment and intelligence. He ordered his endless stream of new recruits to follow simple algorithms, focusing on one (or a few) essential goals. There were always more young Boomers in the pipeline to churn through. The reward function of his process was arguably the most powerful one of all: money.

Microchips need a lot of power, so only a tiny portion of the Earth's mass has been converted so far. But if we keep making more and upgrading them when possible, we won't have to struggle to solve inscrutable problems directly - as long as we pick the right reward function (see: Singularity Soon).

This should arguably be humanity's main goal. It may be necessary to link all the world's microchips into a single "World Mind" that will tackle many different problems in turn.