MT: I thought there were some quite well-established hypotheses (which
doesn't
mean that they're at all right) - by AI thinkers - that sleep might be
essential for, among other things, reprocessing the events of the past
day,
and possibly reprogramming accordingly. >
Jey I've read similar claims about the function of sleep in *humans*. But
why would an AI share such *human* qualities? It's pretty unlikely
that we'll be exactly copying the design of the human brain to build
an AI. After all, we don't have the specific constraints or
"objectives" that evolution had when it "designed" the human brain.
MT: I guess - and do see the detailed discussion Pei referenced - because
the same parts of the brain needed for waking life are also needed for sleep
reprogramming - like the visual system in dreams. This would be all the
more true if AGI/human/animal thought is - and has to be - embodied.That
would mean that if you're thinking about, say, running, then your legs are,
and have to be, involved. (When you think about it, dreams are very physical
affairs, aren't they? There aren't many dreams consisting purely of
intellectual discussions, or logical arguments, are there?)
In principle, of course, you could create an alternative visual system etc.
in your AGI just for reprogramming purposes, but my guess, again, is that
that would be much too complicated as well as wasteful in practice.
-----
This list is sponsored by AGIRI: http://www.agiri.org/email
To unsubscribe or change your options, please go to:
http://v2.listbox.com/member/?member_id=231415&id_secret=28404328-5d2c51
