IMPROVING THE "NEWCONCEPT SCR #9" SUBROUTINE
There have been a few scoffing comments on the Internet after we
posted "fpj28nov98" to Usenet, but today undaunted we are dealing
with the chief obvious glitch, which is garbage showing up in "EX-
AMINE SCR #17" when we try to see which new concepts have been en-
tered into the "lexicon{" array of "NEWCONCEPT SCR #9."
Already we have fixed the problem of making the concept number "n"
show up. We simply make the program print the counter index "I"
to the screen. For some reason, "I" does not even need an inter-
vening "fetch" instruction; it prints directly.
Right now the only concept flag that we are really eager to in-
spect is the final one, the "r" for "recall vector." If the "r"
flag is not being reported properly, then perhaps it is not being
captured properly in the "lexicon{" array.
We need to make the "EXAMINE SCR #17" show only the most recent
three or five concepts, because there is only so much room on the
screen for the human user to view recent input.
Although our attempt at top-down AI in 30K may in fact be doomed
to failure, the smaller minds on the Internet will jeer and scoff
but the superior minds will follow our progress with profound in-
terest. Therefore we state here that the 30K Mind.forth is not
yet up on the Web because we have attacked but not yet gotten rid
of our current glitch. We remind Usenet people that we value our
http://www.scn.org/~mentifex/ theory of mind more than Mind.forth,
which may or may not succeed in embodying the artificial mind.
--
http://www.geocities.com/Athens/Agora/7256/ "The Cyborg Syllabus"
http://www.scn.org/~mentifex/ The Mindmaker Project for Robot AI
