AGI system can be characterized by a way it's implementing certain cognitive
competencies. It might be difficult to grok particular design, but I think that
some of the following issues are central to viable AGI design and probably can
be outlined by short descriptions, which at least give a 'feel' of particular
approach.

It'd be interesting to see short descriptions of existing systems/frameworks
based of following questionnaire.

* How does the system represent structural information (things equivalent to
frame-based representation)?

* How is semantics of information supported in AGI system defined? In
particular,
what is semantics of stored knowledge and of information currently 'active'?
By semantics I mean correspondence to environment through I/O system
or preprogrammed
structure, including time-dependent information.

* How does the system extract patterns in its environment, that is similarities
between separate experiences/world states/objects? How is this
similarity-extraction
related to separate representation of instance of objects sharing
common properties?

* How does concept formation occur?

* How is reasoning/representation dynamics progresses over time and
under influence of
new experience?


-- 
Vladimir Nesov                            mailto:[EMAIL PROTECTED]

-----
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=8660244&id_secret=48014209-f401c0

Reply via email to