On Oct 3, 2008, at 1:40 AM, Rob Freeman wrote:
On Wed, Oct 1, 2008 at 11:40 PM, Christopher Menzel <cmenzel@xxxxxxxx
You have yet to identify a single problem. Can you articulate one?
are you talking about?
I referenced once before here Giuseppe Vitiello's lecture on a basis
for concepts in the dynamics of many-body systems. Perhaps it bears
That you think this work has any direct bearing on AI/OE suggests to
me that perhaps you are under a misimpression of the subject matter of
On what basis do you conflate AI and OE?
VSA's are especially good. They seem to present (compositional)
representation as a vector product in much the way I've been
You do recognize that VSAs are a move in the connectionist AI
paradigm, which of course has many problems and challenges of its own,
right? And that connectionism itself, while a promising approach,
hasn't exactly delivered the goods either, right? And that it's not
clearly even a competitor to the logicist paradigm? You might find
the paper "Is the connectionist-logicist clash one of AI's wonderful
red herrings?" by Selmer Bringsjord worth reading in this regard:
If you looked carefully you would see that Bringsjord's thesis fits
very well with what I have been saying.
Wasn't a reconciliation of the connectionist and logicist paradigms
Reconciling these has been a long-standing goal for many people, and has progressed (about a decade ago) from a research hope to a number of concrete implementations. The most promising, IMO, is the work on using phase-locking in local connectionist models to implement variable binding (which is the key dynamic
aspect of symbolic computation; compositionality is relatively easy). It also has the merit that the neural mechanisms involved have been identified in actual living systems, eg the bat auditory cortex uses this technique to process ultrasound at frequencies higher than neuronal response times.
But in what sense do you think that any of this kind of work represents a critique of logicism, or corrects known 'flaws' in the logical Krep program? It accepts and extends that program, rather than replacing it. A logical ontology implemented on a connectionist engine is no less logical, nor does it have a different semantics, because of the computational architecture it is running on. For myself, I will use any reasoning engine that works, connectionist or not, running on any platform. If someone can make an efficient reasoner using vibrations in crystal pyramids, I'll use it. However, you seemed to be suggesting something much more than a mere engine: you have been referring darkly to 'flaws' and 'pitfalls' in the very idea of using logic at all, no matter what kind of engine it happens to be running on. And you still havn't told us what these awful hidden flaws are.