Self-reflection seems to be a more computationally expensive task than decision making.

The human computer increases its power differently than the accepted mythology suggests. First, the human computer it not an individual brain, it's the entire society with its structure and inter generational transfer of knowledge. The communication between individual "nodes" is part of the structure and the intelligence of the system.

We are largely past the myth of intelligent design, will soon be past the myth of unique sentience, and at some point will realize that the myth of individual exceptionalism is also largely a myth. A genius placed outside of society would be unable to achieve anything.

The way we seem to get past these myths is simply acquiring more computational power through population growth and better communication. Once a certain threshold of computational power is reached a new transformational idea will appear. It's pretty much accepted that if Marconi wasn't around to "invent" the radio, someone else would (indeed, almost every country has a claim to such an inventor). These things are part of the zeitgeist, and it seems that "zeitgeist" is sort of similar to "next token prediction" in many ways.

I am sure that these thoughts will be deeply unpopular, but I am starting to see them more and more. Our working model of the world is shifting.