Scientists have created an adaptive and changeable intelligent material that behaves like the human brain.
One of the greatest challenges of quantum computing is finding materials that allow the next generation of quantum supercomputers to be created. That milestone may be closer today thanks to the work of a team of scientists from Radboud University in Nijmegen, the Netherlands, who have created an adaptive and mutable intelligent material that would be able to behave just like the human brain. The discovery, whose results were published in the authoritative scientific journal Nature Nanotechnology, could contribute to developing the so-called ‘quantum brain’.
Why is this an important discovery?
For Artificial Intelligence to work, computers must be able to recognize patterns and learn new ones. Current supercomputers do this already quite well, but the problem is that they depend on software or a machine learning program that controls the processing of information in hardware. This, however, requires high consumption of energy.
Investigations by researchers from the Radboud University Institute for Molecules and Materials were aimed precisely at implementing machine learning algorithms in hardware based on the combination of different materials and finding out if a hardware component can function without any additional software.
What does this discovery tell us?
Scientists have discovered that this is possible by developing a new intelligent material that learns through physical modifications. In fact, the atoms within the material modify the structure and interconnections between them by mimicking the characteristics of neurons and synapses in the human brain, which, as is known, is not static but works by changing itself physically and chemically.
This atomic-scale machine learning hardware, therefore, not only stores and processes information but also adapts. This mobility was achieved by incorporating cobalt atoms into a black phosphorus structure.
The research began in 2018 when scientists first demonstrated that it is possible to store information in the state of a single cobalt atom. By applying a voltage to the atom, they induced a change of state between a value of 0 and 1, like a neuron.
Subsequently, the technology was developed to create defined structures of atoms, and the most surprising discovery was the ability to vary the connections between atoms as the electrical input varies. By applying long voltages, the synapses changed, and the material adapted accordingly, mimicking the characteristic of biological synapses.
The future implications
This result brings scientists closer to developing a more complex quantum brain, capable of physically adapting and changing according to the learning of neural networks. “If we could eventually build a real machine out of this material, we could build self-learning computing devices that are more energy-efficient and smaller than today’s computers. However, only when we understand how it works, and this is still a mystery, will we be able to regulate its behaviour and start developing it into technology, ” said project chief Alexander Khajetoorians.
It is a known fact that power learning algorithms are highly energy hungry. They consume a lot of energy just to run and function at an optimum level.
With this discovery, we are headed towards a highly technologically advanced computing capacity that is also energy efficient and builds towards a sustainable future. The worldwide demand for computing results in high demand for data centers. This technology could solve the problem of energy footprint while allowing the global demand to be met.
‘It t is clear that we have to find new strategies to store and process information in an energy-efficient way’,
Says Project Leader, Alexander Khajetoorians, Professor Of Scanning probe microscopy at Radboud University.
At its core, the Swiss Institute for Disruptive Innovation believes in changing the world through innovative disruption technology. This discovery is no less than a giant leap forward into having a sustainable future for smarter computers.
Prepare for the future of
quantum computing here
Introduction with practical programming examples