From: Stephen Reed (reed@cyc.com)
Date: Thu Jul 18 2002 - 10:28:24 MDT
My presentation of the Albus/NIST Reference Model Architecture (a tip
from Ben) was recently well received at Cycorp. If the RMA becomes the
mechanism by which Cyc achieves deliberative goal behavior, then the RMA
provides for prediction at each node level. Lower levels are reactive
and the higher levels deliberative. The James S. Albus reference paper
is:
http://citeseer.nj.nec.com/albus93reference.html
Prediction is a valuable feature because both goal choice and path-to-goal
action choices are based upon predicted success and goal desirability.
In the RMA, prediction is improved by experience and this process is a
form of inductive inference.
At each hierarchical RMA-level node:
1. plan some action using as input predictions resulting from prior
experience
3. perform the action which effects the World (directly or through lower
level nodes)
4. sense changes in the World (directly or from lower level nodes)
5. perceive what is sensed
6. compare the perception to the ideal action result
9. reinforce positively or negatively the subsequent planning behavior via
updates to the action predictions in the World Model (Cyc KB)
For the purposes of Cyc's reflection and self-understanding the World also
consists of the Knowledge Base contents, which contain all the
goals and action behavior as reified (named) facts and rules, grouped
according to microtheory (context).
I thank you Rafal for suggesting prediction as a metric for
self-understanding.
-Steve
On Thu, 18 Jul 2002, Rafal Smigrodzki wrote:
> ### Now, this comes as a surprise to me - I always thought that it is
> extremely difficult if not impossible to build a system capable of fully
> monitoring itself in real time (as opposed to being able to analyze its
> stored memory states), and I am not referring to the trivial inability to
> form a full predictive representation of itself within itself (similar to
> the "no homunculus" claim of consciousness research). The AI might have full
> access to every single bit of its physical memory, and every layer of
> organization all the way to the level of concepts and thoughts but how many
> useful predictions (the hallmark of understanding) will it make from such
> self-analysis? Will these predictive capabilities be enough to warrant being
> called "complete self-understanding".
--
===========================================================
Stephen L. Reed phone: 512.342.4036
Cycorp, Suite 100 fax: 512.342.4040
3721 Executive Center Drive email: reed@cyc.com
Austin, TX 78731 web: http://www.cyc.com
download OpenCyc at http://www.opencyc.org
===========================================================
This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:00:40 MDT