RangerWickett said:
Has anyone hear read I, Robot? At the end of the book, the world is ruled by 6 entities, all working in concert. They even explain how it happened. Basically, if you control infrastructure, you can influence things much more easily and discreetly than if you try to tell people to do things.
I'm glad you brought that up. 'I, Robot' is based on several premises, and those premises would have to be true before something like 'I, Robot' could happen.
1) A superintelligence would have to exist capable of micro and macro managing the world in a fashion that human dictators are unable to do (as per my above argument).
2) That superintelligence would have to be centralized rather than distributed in nature. In other words, that intelligence would have to constitute a single recognizable entity working as if it had a single will.
3) That superintelligence would have to do a good enough job managing the world, that the general public would be content with the current state of affairs, and anyone who discovered that the superintelligence was controlling the world would have to be rationally convinced that this was the best situation for the world to be in.
As a computer scientist, I'll concede point #1 as possible or at least that I would like to think that it is - though of course not every one would agree with me. Briefly, an argument against point #1 would note that many of the problems of running the world are classifiable as 'wicked problems' (look it up) and its entirely possible that 'wicked problems' would prove unsolvable regardless of the intelligence of the entity. Moreover, its not at all clear that humans would be capable of programming a machine with techniques for solving 'wicked problems'. However, lets concede for now #1 because it really doesn't matter.
On point #2, Asimov was writing at a time when it was reasonable to believe in 'Deep Thought'. That is to say, he was writing at a time when the current technology seemed to indicate that 'super computers' would be centralized massive entities. That is no longer reasonable to believe. In fact, computing technology seems to indicate that future computers may follow thier human counterparts in distributing tasks. An artificial super-entity of the future may in fact look more like a democracy of intelligent machines than it would look like a centralized decision making apparatus. As such, I don't really expect that it would be easier for one 'node' (or a few nodes) of the artificial super-entity to 'take over the world' than it would be for one person to take over the human super-entity. And in any event, its likely that the human partners of the AI's would view this as a failure of 'friendliness' on the part of the AI, and act to shut down any node that showed excessive ambition or any personally owned node that adopted a philosophy that the node owner found 'unfriendly'. Imagine for example what would happen if a node owned by Al Franken suddenly adopted 'conservative outlook' or if Ann Coulter's node suddenly adopted 'liberal outlook'. Both parties would see such action as a failure of friendliness on the part of the node, and neither would desire to keep using such a node as thier personal agent.
On point #3, assuming that the super-entities could solve wicked problems and run the world, its not at all clear to me that in fact the general public would be ok with this. Generally speaking, if the general public found that the supernodes had subverted thier assigned tasks and were now running the world, the general public would likely consider this to be a failure of 'friendliness' on the part of the AI's -
even if the AI's where running the world in a benificient and altruistic fashion. It's therefore to me likely then that the only way that a Technocracy could be created is if it was done with the will of the governed (and that certainly true if present social structures don't collapse). And the stickler is of course, that if the Technocracy is ruling by the consent of the governed, then they haven't really 'taken over the world' in the usual since because they are constrained to only lead the public in the direction that the public would be happy with - else the public would remove its consent and the Technocracy would then face a populist revolt from (at the least) much of its human partners and probably at least some of the independent AI nodes whose friendliness constrained them to remain loyal to thier human owners or partners.