newbie Mind.Forth Q

Mentifex mentifex at my-deja.com
Wed Aug 30 00:22:05 EST 2000


In article <39AC1629.11BE0F54 at crosswinds.net>,
  Dima Maziuk <dmaziuk at crosswinds.net> wrote:
> Hi all [...]

Mentifex (Arthur T. Murray, author of Mind.Forth):
Hello and thank you for examining my poor attempt at PD AI.

> I've a couple (lots, actually) of questions re Mind.Forth from
> http://www.geocities.com/Athens/Agora/7256/mind4th.html [...]
>
> Q 1: what exactly is ultimate-tag, what does it do?

Mentifex/ATM:
The ultimate-tag variable "ut" is an integer number stored with
any phonemic word in the auditory memory array "ear" and serving
to associate or link the "ultimate" phoneme of each stored word
over to the concept of the word in the lexical array "en(glish)".

When you run Mind.Forth, you may finish a session by typing in:
.ear [RETURN] in order to see (BEC OCMOTP Y HAC) (just kidding)
the final number, or ultimate-tag, near the end of each word-engram.

These numeric tags are a clumsy way of simulating a nerve fiber.

> Q 2: similarly, move-tag?

Mentifex/ATM:
The "move-tag" is a kind of place-holder for an ultimate-tag
when the value is being "moved around" within the software.

> Q 3: how do they work in Comparator? Or, is there a Comparator in
>      pseudocode/plain English that I could have a look at?

Mentifex/ATM:
In the Comparator module of Mind.Forth, the numeric value of
the ultimate-tag "ut" of any recognized input-word is transferred
to the move-tag variable "mt" so that the "tag" can be "moved"
from the old engram of the word to its newest, freshest engram.

On page C4 of the national edition of The New York Times yesterday
(Mon.28.Aug.2000) there was a Human Genome "Code Matching" diagram
which works in a remarkably similar way to the Mind.Forth Comparator
algorithm stemming from the 1993/1994 version of Mind.Rexx AI.
That is to say, the Mind.Forth Comparator recognizes strings of
phonemes not only by matching one unit at a time but also by
"alerting" (NYT term) the next place in line.

The most recent (1.Jul.2000) release of Mind.Forth effortlessy
handles all the various tags, but their function is complicated.

http://www.geocities.com/mentifex/theory5.html is the basic
theory of the underlying mind.model.  Recently I have been sidetracked
a little while I work on a project, The Robot Social Register,
in which I track AI in robots and encourage the use of Mind.Forth.

Please post in news:comp.robotics.misc any and all questions. - ATM
>
> TIA
> Dima [...]

--
http://www.geocities.com/mentifex/bluebook.html#telerobots (q.v.)


Sent via Deja.com http://www.deja.com/
Before you buy.






More information about the Neur-sci mailing list