Chaos/reservoir computing and sequential cognitive models like HTM

This may a digression too far, but one point I thought György Buzsáki was trying to make was the deflation of an ‘object model’ in human vision and thought processing.
The fact that ML is approached by programmers generally, I think makes this problem worse (maybe due to earlier UML and OOP influences).
His view appeared closer to Batesons “the difference that makes a difference” - which is just splitting the data/world anyway that works.
This chimes with your view of living with the contradictions.

1 Like