December 6, 2005
Center of Excellence Proposal
à
[bead thread on curriculum
reform]
The Taos Institute
(on the possibilities)
I feel that the innovations Richard Ballard have developed
are in fact historical... but not easy to understand with the language
used, as in
"conservation laws of nature."
Specifically in the statement
"investors find it too difficult to grasp the concept of an n-dimensional architecture that does not rely on logic and self-consistency. Logic has its place in our technology, but it has a small role to play in comparison to the "conservation laws of nature."
Ballard may correct me if he is tuned in to this discussion, but what is meant here has to be related to the notion of question-answer pairs. It is ultimately a question of empiricism – one has to look at specifics. This is one reason why there is a knowledge builder (editor) in the Mark 3 suite. (If I say anything incorrect, please correct me.)
The legitimacy of a question is a relevant issue. But suppose that one could enumerate all questions about a specific something. This enumeration is the largest space in which to find a conservation of something. Each question-answer pairing is a constraint in an informational space. As Ballard repeatedly reminded me a few years ago, "what is information?" to this his response is "information is the ability to answer questions".
The fundamental principle that Ballard discovered is formal work on localization. His short history is posted at
http://www.ontologystream.com/beads/nationalDebate/262.htm
I am assuming that like other fundamental breakthroughs there are often insights about the breakthrough that is not realized by the people who make the advance.
So as I have listened to Ballard over the years I have come to appreciate that more is going on here that an abstract discussion about matrix inversions.
Specially, I see an invariance over "an ability to answer questions" as one moves from a very large sample of co-occurrence data and collapses the categories – to produce a small conceptual graph (or even Topic Map).
John Sowa talks about this
http://www.jfsowa.com/talks/continuo.pdf
as folding conceptual graphs and spreading activation in conceptual graph constructions.
Something not the same but similar might be occurring as the Latent Semantic Index matrix is iterated, or as stochastic LSI (non-algebraic) processes are iterated. The key here is “categorical collapse” that preserves information (as dimensions are reduced to the smallest number). Does this make sense to anyone? John ? Paul (W)
To paraphrase Ballard: The collapsing of categories is a localization, and this localization is (attempting to) preserve the structure of information, ie to preserve information even as it collapses.
When something like this is said, it is about both the biological phenomenon related to individual mental events (as coherent states of a quantum neuro-wave equation (Pribram)) and about some internal secrete sauce being built (hopefully) into the Mark 3.
We are very early into this, and it is not clear to me where the fault lines are between various contributions.