I actually did a really deep dive into cybernetics, like the actual math involved, which I’m qualified to analyze. It unfortunately would still require a lot of research to get right, and even at the levels of compute we have today it might not be enough. One simply needs to consider the number of types of screws which are necessary to actually fulfill the global demand, and their interchangeability depending on a thousand factors of production, to see the problem. Ultimately, reducing their value to a quantity and optimizing on that quantity based on supply and demand is really easy, compared to some kind of actual graph flow optimization problem based on final product use-demand. The level of democracy at the end would be incredibly complex too, and let’s face it, democracy is not very efficient nor does it even really well reflect the modal persons preferences in a society, let alone representing minority interests.

Ultimately you need a system that can interact with each individuals specific needs and wants (demand), interact with each individual’s abilities, interests, and capacity for labor (supply), even pushing them a bit (incentives), and then balance that interaction with all intermediate necessities to balance the equation, not simply aggregate the averages and expect it to normalize. And even then, think how manipulative, surveillance, and controlling that is.