The development paving the way for next generation computers comes from the U.S. Department of Energy’s (DOE) Argonne National Laboratory and a focal point has been with human forgetfulness. Of interest here is the mechanisms by which our brains discard unnecessary data in order to make space for new information.
To explore this, scientists conducted an experiment that combined supercomputer simulation and X-ray characterization of a new material (quantum perovskite) that gradually “forgets.” This material has the potential to be used for future bio-inspired computing. The biological basis is habituation, which is regarded as one of the primary universal learning mechanisms. This is defined as how an organism responds to repeated stimuli from the external environment.
What is interesting about quantum perovskite is that shows an adaptive response (the potential for a form of ‘habituation’). This occurs when protons are continually inserted and removed. This process of proton addition and removal is said to mimic the brain’s desensitization to a recurring stimulus.
What has been shown by supercomputer simulation and X-ray characterization is that as a proton is added or removed, the material’s atomic structure expands or contracts significantly. This accommodates the proton into a process called ‘lattice breathing’ Here the perovskite’s behavior alters so that the lattice breathing is reduced, causing the electronic properties to evolve with the process.
The addition and subtraction of protons allows the material to resist its electrical current, enabling the perovskite to be effectively programmed, much like a computer. This arises through implementing a learning algorithm generated by the conductance relaxation behavior of a series of perovskites. Overall the performance of perovskite would far exceed that of silicon.
The following video explains more about the process:
In addition, one day perovskite material and the construction of neural network algorithms could lead to the formation of an artificial intelligence capable of reasoning and human-like decision making.
Commenting on the experiments, run at the Argonne Leadership Computing Facility, lead researcher Subramanian Sankaranarayanan told Controlled Environments magazine: “The brain has limited capacity, and it can only function efficiently because it is able to forget.”
He adds: “It’s hard to create a non-living material that shows a pattern resembling a kind of forgetfulness, but the specific material we were working with can actually mimic that kind of behavior.”
The research has been published in the journal Nature Communications, under the study title “Habituation based synaptic plasticity and organismic learning in a quantum perovskite.”
