I finished reading: “Finding an Alternative to Connectionism” which was of particular interest to me as I had been a big proponent of Connectionism since I first heard about it as a teenager in 1990 and watched the growth, decline and rebirth of neural networking in computing.
Results were impressive and yet, their poor model of a “neuron”, which goes back to the late 1940s (and was never intended to be an actual model of human neurons, just a simplification), is what their models rest upon: So, I knew there had to be a better way to do things and I find your model to be superior as it has adaptation as a primitive function rather than assigning weights based upon gambling probabilities.
An interesting find on fast adapting: Look how fast the fruit fly adapts to sound? 10ms. That’s fast. But, flies also can’t do as much as humans and may be at a slightly lower intelligence level overall.
But wow, 10ms is fast adapting.