I barely noticed this paper. It said "Hopfield" but when I looked the references talked about rock salts and excitons. But then I had to look again.
A Fock state pertains to quantum many-body solutions. To get to a Fock state, you need the underlying Hilbert space, which represents "all possible paths". A direct product of two Fock states is exactly that, deriving from all possible paths in an additional dimension.
Fock states are intimately connected with creation and annihilation operators. This is important in human brains because of the time course. A Hopfield network can theoretically operate at the speed of light, even when its neurons have msec refractory periods.
Consider this in relation to Scruffy's model of consciousness as an unfolding of physical time. The key piece of architecture is: the Hopfield network must span the timeline, that way it sees the joint distributions of predictions and results.
Something must "instantiate" prediction. There's no need to predict anything unless something requires it. A universal timeline motif is "voluntary motor behavior". Every muscle contraction is accompanied by a prediction of the sensory consequences. If the sensory input matches the prediction, there is no new information and nothing gets stored in memory. If it doesn't match, something changed and there may be a need for further action (meaning, additional predictions). It is very likely that these sources and destinations of information can be described by creation and annihilation operators.
A Fock state pertains to quantum many-body solutions. To get to a Fock state, you need the underlying Hilbert space, which represents "all possible paths". A direct product of two Fock states is exactly that, deriving from all possible paths in an additional dimension.
Fock states are intimately connected with creation and annihilation operators. This is important in human brains because of the time course. A Hopfield network can theoretically operate at the speed of light, even when its neurons have msec refractory periods.
Consider this in relation to Scruffy's model of consciousness as an unfolding of physical time. The key piece of architecture is: the Hopfield network must span the timeline, that way it sees the joint distributions of predictions and results.
Something must "instantiate" prediction. There's no need to predict anything unless something requires it. A universal timeline motif is "voluntary motor behavior". Every muscle contraction is accompanied by a prediction of the sensory consequences. If the sensory input matches the prediction, there is no new information and nothing gets stored in memory. If it doesn't match, something changed and there may be a need for further action (meaning, additional predictions). It is very likely that these sources and destinations of information can be described by creation and annihilation operators.
