Brain researchers marvel not only at the complexity of the human brain, but also its energy efficiency.  It’s a popular metric in the neuroscience set to say that the brain can literally run itself on two bananas a day.  AI takes orders of magnitude more energy to function, and this is one of the big challenges facing the creation of AGI, or artificial general intelligence.

Artificial neurons.  Now, a team of researchers has designed a way to build a prototype of an artificial neuron that uses ions instead of electrons, mimicking human neurons.  Ions are a key driver of the energy efficiency of electronic transmissions in the human brain.

Memristor effect.  The prototype is made of incredibly thin graphene slits which house a single layer of water molecules, according to a new study published in the journal Science.  The basic unit of the brain is the neuron, and the essence of the brain’s operation is built upon  communication by way of electrical transmissions between neurons.  This new study illustrated the memristor effect, resistors in a circuit that 'remember' their state even if you lose power. 

The researchers were able to create electric fields in the ion channels of the artificial neurons that could assemble a single layer of water molecules into elongated clusters, which developed the memristor effect.

Electronic memory recall.  This landmark work is ongoing with French and British scientists working at the University of Manchester.  The next step for the team is to execute basic learning algorithms which could lay the foundation for electronic memory recall in the artificial neurons.

Easy peasy to share this story with your peeps