🚀 go-pugleaf

User Manual for Mind.Forth Public Domain Artificial Intelligence

[ See also:  Programmer Manual for Mind.Forth PD AI (TBA) ]

Contents by Section:

-- Obtaining Mind.Forth
http://www.geocities.com/Athens/Agora/7256/mind-fpc.html

-- Getting Started (how to run Mind.Forth)

Obtaining Mind.Forth:

WWW "capture" downloads.
CD-ROM's (not available)
PD and Shareware diskettes (not available)
keyboard entry from printed listings (not available)

Getting Started (how to run Mind.Forth)

After setting up an FPC directory on your IBM-clone hard
disk, you may apply the DOS command "rename" to your ASCII
sequential text file of Mind.Forth source code.

The only two requirements for the name of the AI are that
the name be unique on your system (so that it does not
overwrite any pre-existing file) and that it end with the
extension ".SEQ" to identify the AI program as a sequential
file -- as needed by F-PC Forth.  A third consideration is
that the first part of the name (before ".SEQ") ought to
be no longer than eight (8) characters in length, so that
the name of the file will mesh well with any DOS system.

A good choice of name for your own personal AI entity might
be the sort of name that you would (or do) give a robot,
such as "Andru.SEQ" or "Genius.SEQ" or "HeyYou.SEQ".  Above
all, the name should be easy to remember when you want to
bring the AI to life ("Igor.SEQ"?) by typing into the F-PC
screen:  "open Andru.SEQ" and "1 load" and "MIND".

Here is a list of AI entity names for you to consider:

Agent.SEQ Alien.SEQ Angel.SEQ Brain.SEQ Buddy.SEQ Cadet.SEQ
Coach.SEQ Cyborg.SEQ Daddy.SEQ Doctor.SEQ Doorman.SEQ
Dossier.SEQ Driver.SEQ Dude.SEQ Father.SEQ Genius.SEQ
Ghost.SEQ Gofer.SEQ Guard.SEQ Hobo.SEQ Maid.SEQ Mommy.SEQ
Monster.SEQ Mother.SEQ Nurse.SEQ Officer.SEQ Person.SEQ
Robot.SEQ Scrooge.SEQ Sensei.SEQ Sentry.SEQ Servant.SEQ
Shadow.SEQ Slack.SEQ Soldier.SEQ Spy.SEQ Trainee.SEQ
UFO.SEQ Valet.SEQ Vassal.SEQ Waiter.SEQ Watchdog.SEQ
Wizard.SEQ Worker.SEQ Zombie.SEQ

Once you have "Gofer.SEQ" or "Yesman.SEQ" up and running,
the program itself will tell you how to communicate with
the AI and how to attempt to control the AI.  However,
Mind.Forth has special features that you can learn to use
so that you may "grow with the AI," that is, keep current
on new developments in public domain AI and keep up your
level of expertise in operating stand-alone or robot AI.

For instance, the Nine Levels of Diagnostics will take
some getting used to -- since they are subject to change!

But you only need to experiment with the PD AI to learn
some typical settings of how to observe the thinking mind.

Diagnostics Level One is not supposed to show you anything
but your interaction with the AI -- a conversation.
However, if you disengage from the conversation, the AI
should continue with its own internal ruminations and
display or speak them for you to see or hear.

Diagnostics Level Two is supposed to show you which concepts
are active just before the AI thinks a sentence of thought,
so that you may gradually figure out how the AI thinks and
how to predict what the AI is going to say.

Dignostic Levels Three to Five are supposed to show you
only the background computation of the AI in thinking mode,
not in comprehension mode -- which can be very confusing to
you, if you start to type in a sentence and you see the AI
begin to analyse your input long before you are finished.
Please note:  There is no back-space function in the AI,
and so you may not correct yourself if you make a mistake.
The AI program needs all the time it can get to process
your words, so it does not have the option of retroaction.
However, if you consistently misspell the same word in the
same wrong way all the time, the AI will accept your word.

Diagnostic levels Five to Nine are supposed to show you
more and more background computation of BOTH user input
AND robot response, but with a "dot-quote" sum-up feature
that shows you the entire string of AI verbal output AFTER
the individual words have raced past you in a blurry scroll.

On one of the diagnostic option levels, you will be able to
set the AI to show its most impressive background thinking
processes ceaselessly until countermanded, as a display mode
for use in countless situations, such as:

- computer fairs
- store windows
- partytime activity
- cable TV dead time (quasi test pattern)
- AI screen saver
- educational settings
- information kiosk
- expert system advisor
- intelligent digester of incoming messages
- intelligence snoop of passing-by messages
- Orwellian monitor of all that goes on
- reader of stories with stop-and-discuss mode
- message announcer ("You've got memes")
- receptionist/sentry

Until eternal loop-around is programmed in, your AI will
come up against its "timed-out" limit and stop running
because its memory channels have filled up.  At that time,
you can still inspect the various memory arrays by entering:

.psi -- to see the innermost mindcore concepts and their tags;

.uk -- to see the contents of the English lexical array;

.ear -- to see the engrams in the auditory memory channel.

If you reenter the main AI Forth word "MIND", the AI will
erase all its memory banks and begin a new session with you,
starting all over again with the time counter reset to zero.

Because this user manual for Mind.Forth and any other PD AI
programs is in a state of rapid change and development, you
may contribute your ideas into the public domain of AI by
any channel, such as posting a message in a Usenet forum.

Because the AI user manual is in the public domain, you may
modify it with your own information, print it and sell it.
The whole purpose of the PD AI and its documentation is to
usher in the age of intelligent machines from the year 2000.