Consider a future human society scenario where a 'supercomputer' was entrusted by all of humanity to divy out the jobs and tasks of all the citizens on Earth. The citizens all put faith in the correctness of each and every decision the supercomputer made, ranging from the grandest of humanity's goals to the smallest minutia of daily chores. And they are right to do so, because it turns out that the supercomputer indeed does manage the earth's resources and as well, the interrelationships of individual humans and their wants and needs, much better than in fact the humans can themselves. This supercomputer would be not officially an AI, but its programming input would come directly from the individual citizens. Each citizen would have a sensor/transmitter on his body that monitored all life functions and senses felt by the citizen. These in turn would be transmitted to the supercomputer who would run a series of algorithms on the information and decide the best route for everyone to take. The algorithms would themselves be programmed by each individual citizen as they go about their daily lives naturally finding improvements and efficiencies in life and industry. We humans would update the supercomputer like some wikipedia entry and each update would hyperlink to many others, thus creating some hyperplexus of information. It would be nice to think, 'Oh, I need some milk, I better go to the store.' And twenty seconds later a neighbour rings your bell and delivers a 2L, because the supercomputer told him that, since your neighbour was already at the store and planning on going home, that it is a more efficient decision for humanity in general and for the wants and needs of the individual, if he picks up the milk and delivers it to you. Of course this means you must consider the flip side of the coin which is that you will be called upon to add random tasks to your daily routine, but I think that would be quite an amusing life and it certainly holds the charm of a well natured spontaneity.