> -----Original Message-----
> From: Nanograte Knowledge Technologies via AGI [mailto:[email protected]]
> 
> How about we discuss how to build an adaptive, complex machine?
> 
> Proposed definition:
> An adaptive, complex machine would have the ability to learn from its
> installable knowledge base and environmental experiences and be able to
> consistently apply functional evidence of this learning in order to solve
ever-
> complex problems with in a fraction of human-effort time.

I picture many agents, thinking about bacteria lately, not sure how many
agents are needed to reach a particular level of adaptiveness. Some systems
might require three agents some a billion. It depends on things,
computational resources and, complexity of agents including individual
communication complexity characteristics and how the complexity scales and
forms patterns across groups of agents IOW how the complexity and entropy of
the MAS behaves - perhaps how adaptive patterns are formed? Interesting
looking at cymatics here I wonder if in a massive MAS, MMAS? if there is
something similar going on since you have this inverse-quantization effect
forming wave signals and the agents can be coordinated to focus waveforms in
various ways to generate computational patterns... or - internally let's say
in biological cells the DNA does perform quantum networking/coordination -
similar could be attempted in a computer MAS cymatically.

I would say engineer into each agent some of these preconceived limiting
unknowns in how the system behaves. For example: How many agents would the
system require to adapt? We don't know so engineer into each the ability to
control how the whole system scales. And then have feedback into each agent
from the whole system - sounds like a genetic system and it could have that
adaptiveness but - keep in mind we are in a computer and have different
computational characteristics verses biologic molecular. Comparably there
are impediments and enhancements but we needn't reverse engineer the
limitations in molecular biology such as in artificial life that's where
that goes very wrong I think. Also ALife intelligence computation might have
gone wrong in being overly focused on genetics in terms of evolution and not
on DNA quantum communication abilities... 

John






-------------------------------------------
AGI
Archives: https://www.listbox.com/member/archive/303/=now
RSS Feed: https://www.listbox.com/member/archive/rss/303/21088071-f452e424
Modify Your Subscription: 
https://www.listbox.com/member/?member_id=21088071&id_secret=21088071-58d57657
Powered by Listbox: http://www.listbox.com

Reply via email to