On Sun, 31 Oct 1999, Robert Owen wrote:
> Has anyone considered the hypothesis suggested by sleep research
> and Hamlet that we sleep in order to dream?
This fits with my hypothesis in that dreaming may be part of the processing of the days short term memories and/or reflex reaction behavior weighting. You could break it down into components that might conform to the various stages of sleep -- memory transfer, reality integration, neural net weighting, behavior rehersal for integration.
I had a dream this morning arguing where I was explaining to several astronomers whom I had never met that the gravitational microlensing observations were consistent with possible designs for SIs and "immortal" civilizations. Now the interesting thing is that when I'm awake I have *no* pictographic memory, i.e. I can't recall (with precise clarity) what a face or my car or my house looks like. I can do something like walk around them. So awake my spatial sense or procedural navigation skill seems to dominate. The interesting thing about the dream was that I had very clear visions of the astronomers' faces and knew that I recognized one but not the other. So in dreams I can access skills that I have (facial recall) but am unable to control at a conscious level.
I'm guessing that my belief system & response system has some heavy weights on "supercomputers in space". :-)
> Now, if we knew  how to induce a transient conscious analog
> of dreaming that would perform the required function when awake,
I believe this is what happens with the micro-"sleeps" that occur with sleep deprivation. But during those micro-sleeps you are probably "offline" with regard to reality so this may not be what you want (e.g. its really bad when driving a car).
> or  how to modify cognition and/or conation so that e.g. there
> was no residuum requiring reduction by dreaming, we might be in
> a position to dispense with sleep per se.
I think we need more processing power or circuits that aren't built into the brain at this point. As, I think Anders pointed out, we had lots of down time at night during our evolution (or during the day for nocturnal creatures) in which to develop these processing methods. They are probably tied in at the lowest level of brain structure and are going to be difficult to re-engineer.