You do realize though that the concept of optimization of an utility function is a very stylized working model for rational decision taking, and not necessarily a hi-fi representation of how humans take their decisions in practice.
Humans do not have utility functions and optimizing algorithms running online in their heads.
They have hardwired response to stimuli and adaptive behaviors, which, after complex interactions with an environment, do sometimes behave as if there were implicit utility functions being optimized, and that’s why the model work, to some extent.
But utility function profiles do change with circumstances, and these interactions are very complex and we are still far away from creating dynamic models that capture all these features. (see also my post about rational expectations: https://forum.freecapitalists.org/t/mathematical-expectation-and-rational-decision-taking-split-from-defense-in-anarchy-thread/26282 )
This is more to the point of what I was calling simulating the human brain
The brain process sensorial information and also information from other channels physiological channels beyond those we call “senses”.
Once we’ve understood the micro and macro structure of the brain, and how it grows and adapts to stimuli, it would be possible to envision a complex enough system that respond to these data inputs in a likewise fashion, but running on silicon.
Sure, the concept of creating a simulation of an entire brain is right now of speculative, maybe academic, interest.
For most applications we just need some functionalities, like vision processing and etc.
Those are “simpler” problems that are already hard enough to occupy lots of scientists, and for which applications seem more realistic in a near future.
But once enough knowledge about the parts is in hand, it becomes increasingly interesting to look after the whole thing.