Throughout history, artificial intelligence has been a hot topic: A computer algorithm "learns" by being taught examples of What is "right" and what is "wrong." But neurons are the brain's cells, not computer algorithms.

Signals are sent from these neurons to other neurons. This complex network of neurons and synapses controls our thoughts and actions.

Biological signals

In comparison with conventional computers, biological signals are much more diverse. Neurotransmitters, ions, and biomolecules communicate among neurons in a physical neural network.

In particular, neurons communicate chemically by emitting messenger substances, such as neurotransmitters, or electrically by sending electrical impulses called "action potentials" or "spikes." 

Research is currently being conducted on artificial neurons. To achieve efficient communication between biology and electronics, artificial neurons must realistically emulate the functions of their biological counterparts.

In other words, artificial neurons can process a variety of physical signals. Most artificial neurons have emulated their natural counterparts electrically without considering the wet natural environment containing ions, biomolecules, and neurotransmitters.

The Max Planck Institute for Polymer Research, led by Paschalis Gkoupidenis, has now developed the first bio-realistic artificial neuron by tackling this problem.

As a result, this neuron can produce spiking dynamics similar to those found in biology and can thus communicate with "real" biological counterparts. 

The group of Gkoupidenis developed a nonlinear element made of soft organic matter, also found in biological neurons.

This enabled researchers to create a realistic artificial neuron capable of communicating in a biological environment in various ways, such as chemically or using ionic charge carriers.

Source: https://techxplore.com/news/2022-11-language-artificial-neurons-mimic-biological.html