> A simple question in reply to this.
> Why does perception, or any other aspect of AGI, have to follow
> the model of
> what natural systems (like human beings) do?
> Could not an AGI system simply be different, built for different
> assumptions
> and modeling?
> We do know that natural systems work, but we also know that they
> have flaws.
> We don't know either thing about other models; whe simply have to
> find out.
> C. David Noziglia


I think most of us on this list agree with the feeling you're expressing,

My own Novamente AGI project is not based on closely modeling the human

However, the human brain/mind is the best example of a generally intelligent
system that we have today.  So it certainly makes sense to pay some
attention to how it works!

In our Novamente work, we have drifted further and further away from
human-brain/mind-based thinking and designing, as the project has

The first version of Webmind, circa 1997 [Webmind, 1997-2001, was the
predecessor system to Novamente] was much closer to the human brain/mind in
concept than Novamente [our current AGI design, fully designed but far from
fully implemented].

Analogy to the human brain/mind helped us to "prop up" our ideas in the
early stages, but once we really felt we understood what structures and
dynamics were needed for AGI, we began thinking more and more about how to
achieve them efficiently on clusters of SMP machines, and how to best
exploit the particular capabilities of contemporary commodity hardware.

One of the problems that computer scientists seem to have when creating
would-be AGI systems *without* paying close attention to the human
mind/brain, is oversimplification.  Computer scientists seem to gravitate
toward simple, elegant models that at best capture one small aspect of
general intelligence.  Attention to the human mind/brain is valuable, among
other regards, in that it suggests the complexity, diversity and "messiness"
that *may* be necessary in order to achieve significant AGI under highly
limited space/time resources.

-- Ben G

To unsubscribe, change your address, or temporarily deactivate your subscription, 
please go to http://v2.listbox.com/member/

Reply via email to