M
mentifex
Java knowledge is a major job skill for programmers.
AI knowledge is not yet much in demand, but it will
be once word gets out that AI has been solved and
needs to be ported from Forth into Java for robots.
Since the AI software solution on 22 January 2008,
we are improving the basic, original AI Mind and we
are releasing our AI "labnotes" to spread the word.
Today we are tweaking the "EEG" AI flatliner code
that brings a brain-dead AI back to life when the
artificial intelligence has gotten stuck in a rut by
generating the same thought repeatedly or by
failing to generate any thought at all.
The artificial mind, whether executed in Forth or
Java, needs to reactivate its old concepts when a
human user fails to answer a question posed by the
AI about some new concept (or typographical error)
entered by the user during dialog with the AI Mind.
Previously we have gotten the AI to think of the
oldest concept in its knowledge base (KB); now we
need to engineer not just remembering the concept
but also creating a thought about the concept,
so that the robot AI Mind will let its musings
wander along a meandering chain of thought.
Since we want high school teachers and members of
robotics clubs not only to demonstrate the AI but
also to see how long they can keep the AI alive on
a computer surrounded by unpredictable human beings,
we need to keep the meandering chains of thought
going so that human beings will be sufficiently
intrigued to devote time in service of the AI Mind.
Accordingly we open up the codaver for psychosurgery.
In the display of output below, we see that the oldest
concept "CATS" is briefly the reentrant "psi" concept
in Audition. Therefore we need to trap that "psi" value
as the "redux" concept to be brought back to life.
Calling EGO (AI4U Chapter 20); inert = 15
Repetitious thought detected; EGO module steps in.
EGO Module recalls the oldest memory --
C
AUD: psi topic redux 0 82 82 A
AUD: psi topic redux 0 82 82 T
AUD: psi topic redux 0 82 82 S
AUD: psi topic redux 76 82 82
AUD: psi topic redux 0 82 82
EGO: psi & topic = 0 82
English calls Ask to increase the AI knowledge base.
Ask-module calls whatAuxSDo to generate a question.
After we trap the Audition "psi" value only during
rentry and only when "EEG" is at a flatliner zero,
we see the AI resurrect the oldest concept #76 "CATS".
AUD: EEG = 1 psi = 0 topic = 82 redux = 0
SVO: thotcyc & thtnum = 3 82000000
SVO: thot1 = 82000000
SVO: thot2 = 82000000
SVO: thot3 = 82000000
Calling EGO (AI4U Chapter 20); inert = 15
Repetitious thought detected; EGO module steps in.
EGO Module recalls the oldest memory --
C
AUD: EEG = 0 psi = 0 topic = 82 redux = 0 A
AUD: EEG = 0 psi = 0 topic = 82 redux = 0 T
AUD: EEG = 0 psi = 0 topic = 82 redux = 0 S
AUD: EEG = 0 psi = 76 topic = 82 redux = 76
AUD: EEG = 0 psi = 0 topic = 82 redux = 76
EGO: EEG = 0 psi = 0 topic = 82 redux = 76
Finally, the oldest concept springs to life
as the subject of a sentence, so that the AI
may follow meandering chains of thought.
Now who will code the AI Mind in Java?
ATM/Mentifex
AI knowledge is not yet much in demand, but it will
be once word gets out that AI has been solved and
needs to be ported from Forth into Java for robots.
Since the AI software solution on 22 January 2008,
we are improving the basic, original AI Mind and we
are releasing our AI "labnotes" to spread the word.
Today we are tweaking the "EEG" AI flatliner code
that brings a brain-dead AI back to life when the
artificial intelligence has gotten stuck in a rut by
generating the same thought repeatedly or by
failing to generate any thought at all.
The artificial mind, whether executed in Forth or
Java, needs to reactivate its old concepts when a
human user fails to answer a question posed by the
AI about some new concept (or typographical error)
entered by the user during dialog with the AI Mind.
Previously we have gotten the AI to think of the
oldest concept in its knowledge base (KB); now we
need to engineer not just remembering the concept
but also creating a thought about the concept,
so that the robot AI Mind will let its musings
wander along a meandering chain of thought.
Since we want high school teachers and members of
robotics clubs not only to demonstrate the AI but
also to see how long they can keep the AI alive on
a computer surrounded by unpredictable human beings,
we need to keep the meandering chains of thought
going so that human beings will be sufficiently
intrigued to devote time in service of the AI Mind.
Accordingly we open up the codaver for psychosurgery.
In the display of output below, we see that the oldest
concept "CATS" is briefly the reentrant "psi" concept
in Audition. Therefore we need to trap that "psi" value
as the "redux" concept to be brought back to life.
Calling EGO (AI4U Chapter 20); inert = 15
Repetitious thought detected; EGO module steps in.
EGO Module recalls the oldest memory --
C
AUD: psi topic redux 0 82 82 A
AUD: psi topic redux 0 82 82 T
AUD: psi topic redux 0 82 82 S
AUD: psi topic redux 76 82 82
AUD: psi topic redux 0 82 82
EGO: psi & topic = 0 82
English calls Ask to increase the AI knowledge base.
Ask-module calls whatAuxSDo to generate a question.
After we trap the Audition "psi" value only during
rentry and only when "EEG" is at a flatliner zero,
we see the AI resurrect the oldest concept #76 "CATS".
AUD: EEG = 1 psi = 0 topic = 82 redux = 0
SVO: thotcyc & thtnum = 3 82000000
SVO: thot1 = 82000000
SVO: thot2 = 82000000
SVO: thot3 = 82000000
Calling EGO (AI4U Chapter 20); inert = 15
Repetitious thought detected; EGO module steps in.
EGO Module recalls the oldest memory --
C
AUD: EEG = 0 psi = 0 topic = 82 redux = 0 A
AUD: EEG = 0 psi = 0 topic = 82 redux = 0 T
AUD: EEG = 0 psi = 0 topic = 82 redux = 0 S
AUD: EEG = 0 psi = 76 topic = 82 redux = 76
AUD: EEG = 0 psi = 0 topic = 82 redux = 76
EGO: EEG = 0 psi = 0 topic = 82 redux = 76
Finally, the oldest concept springs to life
as the subject of a sentence, so that the AI
may follow meandering chains of thought.
Now who will code the AI Mind in Java?
ATM/Mentifex