[Avery Andrews 931130.1249]
On the subject of feedforward diagrams, I like mine better than Hans',
since Hans' seems to imply that there is a single coherent world-model
whose-state is being controlled, which seems to me to be rather dubious
(c.f. Dennett's _Consciousness Explained_). Mine is supposed to be
saying that any particular control system might (or might not) have
a model-assist (I think model-assisted is a good term, since it it clear
that the models can only play a supporting role, and not for very long),
without making any claim that these cohere into a single model of the
world.
Avery.Andrews@anu.edu.au