Content-type: text/html Downes.ca ~ Stephen's Web ~ To create living AI, replace neural networks with neural matrices

Stephen Downes

Knowledge, Learning, Community

I have no idea how to evaluate the argument made in this paper, but it's definitely an interesting perspective. Alex Kostikov notes that "the basis for the plasticity of living neural networks is not the structure of the network as a set of synapses and neurons, but a special class of transmembrane proteins that form ion channels on the surface of each individual neuron." These allow neurons to be "carriers of their own individuality, capable of changing their own attitude to the incoming signal in real time... it shows how a matrix of preferences is formed from billions of individually special neurons." The question is, are there patterns in this matrix of preferences that we can describe, as Kostikov seems to suggest, or is any such matrix fundamentally ineffable by virtue of its complexity?

Today: 0 Total: 8 [Direct link] [Share]


Stephen Downes Stephen Downes, Casselman, Canada
stephen@downes.ca

Copyright 2024
Last Updated: May 30, 2024 03:35 a.m.

Canadian Flag Creative Commons License.

Force:yes