Charles H:I'm not sure that "Therefore we need a bunch of AGIs rather than
one"
is a valid conclusion, but it *is* a defensible one. I have a suspicion
that a part of the reason for the success of humans as problem solvers
is that not only do we start from a large number of initial positions,
but also we tend to have differing goals, so when one is blocked,
another may not be. But I see no reason why, intrinsically, the same
characteristics couldn't be given to a single AGI...though the variation
in goals might be difficult.
Well put. (BTW as perspective here, I should point out that what I've raised
calls for a whole new branch/dimension of social psychology - the study of
collective intelligence. We really have a v. poor to nonexistent
understanding of how that works. We have a v. poor picture for example of
how science (i.e. groups of scientists) builds up its collective pictures/
bodies of knowledge. What we are discussing here are some of the possible
elements of a new science. It would also cover collective animal
intelligence such as the extraordinary collective activities of termites).
But your valiant defense of a superAGI possibly splitting itself up into
parts to mimic a society won't work - as you immediately start to realise.
Any individual agent must establish sets of priorities among its goals- and
among its various activities (and even "subselves"), because they are all
competing for its limited time, energies, and resources. Different
individuals can establish totally different priorities. And each human
individual does indeed have an (often frustratingly) different POV and
philosophy to one's own (which is of course always the right one). So a
whole set of separate intelligences can be aware of far more factors/
evidence, contribute far more different personal experiences, generate far
more ideas and subdivide far more labour than an individual intelligence.
(What's the popular science work that came out recently about the
superiority of collective intelligence?)
Another thought here is that language and maths - which would surely be
vital for a superAGI - probably only evolved when you had the beginnings of
truly complex social activities - where you didn't just have, say, group
hunting (which animals of course do) but hunting with sets of tools and
maps, pictures and planning, and, similarly, complex construction
activities - where you needed complex coordination of individuals.Language
and math IOW exist to finely coordinate separate minds.
-----
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=4007604&id_secret=59181610-de9996