ordo30may2KA -- sequence of ratiocination in Amiga Mind.Forth --
http://www.geocities.com/mentifex/mind4th.html -- Wintel version.
#38-ALIFE calls #4-PSI-CLEAR immediately prior to #20-SENSORIUM,
so that all `psi' from current `t' backwards are set to zero.
#20-SENSORIUM orchestrates input of up to three words.
If a new word is entered,
#14-NEWCONCEPT assigns to it an activation of about ____ 30.
If an old word is entered,
#13-OLDCONCEPT assigns to it an activation of about ____ 30 and
calls #10-ACTIVATE to reactivate old nodes of the same concept.
Then #10 ACTIVATE calls #9-SPREADACT.
#9-SPREADACT transfers `bulge' to `pre' and `seq'.
#38-ALIFE calls #35-THINK (TRANSFORMATION) to generate a sentence.
#35-THINK forces a verb to become active, by calling
#10-ACTIVATE (which further invokes SPREADACT) and
#34-DISCRIMINATE (which finds the most active concept).
#35-THINK calls the #33-ENGLISH module to think a thought,
which re-enters the mind word-by-word into #20-SENSORIUM.
#38-ALIFE again calls #4-PSI-CLEAR to repeat the above process.
Tues.30.May.2000 Usenet note: Mind.Forth PD AI is on the home
stretch of becoming an AI suitable for installation on robots.
The above "ordo cogitandi" document is being used to fine-tune
the process of `spreading activation' so that Mind.Forth will
generate only logically true statements and not talk nonsense.
The document is being posted here as a progress report and as
in invitation to be present in Cyberspace at the dawn of AI.
--
Forth is the lingua franca of AI, embedded NC and robotics.
For details on Mind.Forth and the public domain AI project,
see the ACM Sigplan Notices 33(12):25-31 (December 1998),
"Mind.Forth: Thoughts on Artificial Intelligence and Forth."
