|
Another vote for omnibenevolent ai here. The problem with benevolent dictatorship is that it's not a plan so much as a wish. How do you get that nice, smart person into power and keep him there? What happens when he croaks? Rome tried this at various times and its just not stable. Besides you run into all the usual issues concerning human corruptability anyway. Program an ai to govern. Give it a sarcastic sense of humor so people dont feel alienated, and to keep things interesting. Wahlah. Perfect governance and religion too if you want to play it that way.
|
#
¿
Sep 23, 2015 21:38
|
|
|
|
| # ¿ May 15, 2024 07:16 |
|
|
LookingGodIntheEye posted:I don't know why we assume a machine made by humans is going to be free from human flaws, or that its morality will be compatible with ours without essentially encoding our current forms of governance (and all its problems) into a more efficient, easily enforced and more entrenched format. It's not a given that a machine would be incorruptable, but they follow the rules they are given much more discretely than a human is able to. Even "learning" algorithms are incapable of going completely off the rails of their programming. This is something i dont expect to change even as we build more true intelligence into our systems. Artificial evolution can produce novel configurations that might not be expected but they still perform the tasks the original iteration was engineered to accomplish. We're not talking about the difference between an automaton and a true intelligence either. We're talking about a difference that is steeped in the physical differences between biochemical and electrical "thought". IMO, at least. I could be completely wrong about it. It might be that we get machine self programming to a level that negates those core differences. Either way, I agree that incompatibility might well become an issue. Even perfect omnibenevolence might seem horrific at human eye level and people would resent the gently caress out of being governed by a godlike computer no matter how manifestly awesome it was. This is where the humor comes in. I think it's utterly crucial. It's not just an ian banks reference. How do you ensure a software deity remains human enough to be acceptable? Make it sarcastic as hell, obviously. The other issue that seems obvious to me is what rudatron touched on, that power is a process and it is not held in a vacuum. In this case, the computer needs maintence and probablly additional programming occasionally. The people that do this are the de facto watchers of the watchman and the highest rung of power. Whoops, we're back to oligarchy. I can't think any good way around this, excepting a magically and scarily self sufficient machine. If the computer doesn't need people at all you're walking into the setup of an alarmist sci fi movie though.
|
|
#
¿
Sep 24, 2015 07:29
|
|
|
Chucat posted:Can we give Athenian Democracy another try, that was pretty fun (but let women in). With the tech we have it might actually be feasible to construct that kind of direct voting, everyone yelling at each other kind of democracy but i still wouldn't want it. I flat out don't trust the public with that kind of power. And not just to set myself up as agreeing with some of the smarter founding fathers. I've seen public stupidity. It's not that humans are individually idiotic, we're just collectively idiotic. Mobs are one of the scariest things on the planet and i certainly don't want to be ruled by the passions of one.
|
|
#
¿
Sep 26, 2015 17:19
|
|