__________ ___________ / \ ___________ / \ ( TABULARASA ) / \ / motorium \ \__________/ / BOOTSTRAP \ \_____________/\ __|___ /\_____________/ __________ \ / \ / ________ / \ \/ MAIN \/ / \ ( volition )--------< ALIFE >----------( SECURITY ) \__________/ /\ LOOP /\ \________/ _____________ / \______/ \ _____________ | / \/ | \/ \ | _______ \ THINK / ____V_____ \ SENSORIUM / | / \/\___________/ / \ \___________/ _V_ ( ENGLISH ) ___________ ( emotion ) | / \ \_______/ / \ \__________/ | ( HCI ) | ( VERB-PHRASE ) ________ | \___/ | /\___________/\ / \ ___V____________ __|__ / | | \/ \ / \ / \/ __V__ | ( SPEECH )--( AUDITION ) ( SVO ) / \ | \ / /\________________/ \_____/ ( REIFY ) | \________/ / | | | \_____/ | / / | | | | | / / __V__ _V___ V__________|______V / ______ / / \ / \ / \/ / \/ / OLD- \ / NEW- \ \ NOUN-PHRASE / ( audSTM ) ( CONCEPT ) ( CONCEPT ) \_________________/ \______/ \ / \ / / /\_____/ \_____/ ________ / ________/ | \_______/ | / \/ / \ | / \ | ( audRecog ) ( ACTIVATE ) | ( enVocab ) | \________/ \________/ | \_______/ | _________ / V_____________V / \/ / \ ( SPREADACT ) ( INSTANTIATE ) \_________/ \_______________/
As discussed on
KurzweilAI.net Mind-eXchange, the expanded flowchart or
diagram above provides an overview essential for understanding the AI Mind
well enough to
code a new version of the AI in a new programming language.
The diagram shows not only the modules called by the main
alife loop (q.v.),
but also the submodules called in the course of the mental processes of the AI.
Some modules, such as Tabularasa, may be necessary in one language (e.g.,
Forth) but not necessary in other programming languages (e.g., JavaScript).
A given module may be called only once, as at start-up when Bootstrap
is called,
or only when necessary, as for example when Security
is called to handle an event.
Some of the modules shown in lower case (e.g., volition
and motorium)
have been prepared in theory but have not yet been implemented in code
and are shown here mainly to indicate their future location in the AI Mind.
The clockwise sequence of the modules in the main alife loop at the top
of the diagram is not haphazard and has been designed very carefully.
Each module is located after any necessary prerequisite and
before any module depending on the results of the current operation.
Thus for instance Sensorium comes before
Think so that a robot
may first have a sensory experience of, say, user input from a human
and then think up a spoken response to the input from homo sapiens.
After and in reaction to the Sensorium,
the emotion module engages
the embodied mind so as immediately to influence the
Think module.
The Think module seeks truth before volition
because free will
is logically based upon the verisimilitude achieved by thinking.
Then volition orders the robotic action of the
motorium.
Note that the AI flowchart shows only how one mind module
calls another, and not the underlying flow of information and
reification that results when a module modifies information in
memory or passes data from one memory channel to another.
For instance, the Speech module and the
audSTM module
(auditory Short Term Memory) are not directly connected above
but they both process data stored in the auditory memory array.
In the flowchart above, the mechanisms of control are more
important than the pathways of data-flow -- which are shown not
here but in other diagrams associated with the Theory of Mind.
The arrangements in the above diagram are only a starting point
for the creation of artificial minds. Anyone may alter the modular
relationships in a divergently evolving species of the Robot Mind.
Eventually there will be a much different schematic when there are
massively parallel implementations of the artificial intelligence.
For the survival of the fittest mind design we must pass through
the imperfect and temporary lay-out of the flowchart above.
/projects/mind
Last updated: 3 September 2003
Return to top; or to the
weblog.html Weblog-Sitemap.