Nit, the 2nd: Kevin [hope not too informal] uses the eg. of an internal signal generated by a cell surface receptor in chemotaxis as a signal that is not a representation. I suggest instead, this signal = degenerate (simplest) form of rep. It’s an arbitrary info. vehicle that can rep. only 1 thing
[Ack! Was gonna complain you mention “semantic pointers” w/o referencing Chris Eliasmith’s Semantic Pointer Architecture, but I see you referenced the Blouw paper, which was Eliasmith’s. BTW, you need to add a line space in front of that paper in the ref’s at bottom.] @wiringthebrain.bsky.social