I will describe the recent work we are pursuing in a new field that I call Solid State Neuroscience. We are building individual electronic neurons based on a novel memristive system, which is made from conventional, off-the-shelf, electronic components.
Moreover, our neurons are of unprecedented simplicity, have electric spiking behavior, can be interconnected as Lego blocks to build networks, and naturally work on timescales compatible with biology. This can be considered as an experimental platform where to test mathematical models, to implement algorithms for artificial intelligence (AI) applications, and, ambitiously, to interact with living animals to build Brain-Machine-Interfaces. One may also think of our methodology as a concrete realization of a Neural Simulator to study neural networks, analogous to the idea of a Quantum Simulator to study quantum models, as proposed by R. Feynman.
