Pay attention. This video is from 1992.
Error 404 (Not Found)!!1
Sorry, catching up here.
The biggest reason the Soviet planning didn't work?
In two words, Human Nature.
Even as individuals we display random seeming (to an outside observer, ad sometimes to ourselves) behaviour, desires, urges to do, learn, create, various things which are outside our ordinary day to day pattern of behaviour.
It might be something as simple as seeing what's behind that rock over there, taking a different route just for the heck of it, or diving into a new subject that strikes our fancy. While sweeping generalizations can be made in retrospect, simply put, humans are not completely predictable and the Devil is in the details. Who would have known some guy importing beach pebbles from Mexico by the dump truck load and sticking them in little cages with an inexpensive but cute pamphlet would become a millionaire selling "Pet Rocks"? Or there would be fights in supermarkets over "Tickle Me Elmo"?
While much of that is marketing, something which appeals often not to the rational nature of people, but which is designed to make people want things they often do not
need, our wants ultimately drive us to do extraordinary things, above and beyond the relatively mundane basics. Not only that, but those 'basics' are different, depending on where and
how you live, that commonly the ultimate reflection of your personal wants and needs.
Three of the six vehicles in the 'fleet' are 4wd capable, one an all wheel drive, the others 2WD, but all have tires far more aggressive than those which might be seen as practical in a more temperate and urban climate, all of necessity. (Not every day necessity, but for those times when you absolutely have to get somewhere and nothing less will do.) Where I grew up, those would be seen as 'overkill', but here, they are rather tame examples compared to the common pickups with lift kits, oversized tires and wheels, and other modifications: mine are pretty much factory stock.
No central planning department in a nation as broad and varied as this one will succeed, because it is not in human nature to have identical wants or needs, outside of that for fundamental survival: food, clothing, water, shelter, breathable air, at a temperature we can survive or with the means to alter that to a level which supports human life (up or down).
The successful use of an AI to interface with humans would have to be capable of dealing not just with the rational things humans do, but the irrational, the impulsive, the tweaks that occur in a daily routine from curiosity, spotting or noticing something new, and the rational improbabilities of being delayed by stopping to help someone, having an urge to see an old friend, or any of the myriad other things that slow a human down, right down to slowing down or stopping to admire a particularly beautiful sight. Human motivation may sometimes be predictable, but not always, and there is the bugaboo.
The other issue, and one that will ever remain with any attempt to quantify human behaviour is that not all answers will be honest or complete, adding yet another level of complexity to the equations.
With data which are based on generalizations or questionable honesty, garbage in, garbage out. Every AI would have to develop its own personality. Which means that with even the most careful handling, some of them just won't be wired right...the fodder for horror movies, or plot twists from the HAL 9000 on.
As to the original question, at the start of this all, well, it would depend on the human or the machine. I'd be biased in favor of keeping the human, generally, because of a species bias, not to mention the Commandment against murder, but that begs the question of whether a truly sentient machine would have a soul.
And that circles back to our learned responses to the idea of pain or mortality, are we thinking any more than a machine programmed to react the same way? Or do our proclaimed instincts go even deeper?
Good stuff Fred, thanks.