A new demo of a brain-inspired device, called a memristor, took place recently. It was meant to power artificial systems. It can mimic the human brain.
A nanoscale gadget termed a memristor could be employed to fuel artificial systems that copy the human brain. These artificial neural networks (ANNs) show the ability to learn new stuff and they can also do things that are a huge burden for ordinary run-of-the-mill computers to accomplish.
Don't Miss: Today's Best Deals on Amazon.com
These include: pattern discernment, online learning and slotting information. Solid ANN policies are hard to implement due to a total lack of efficient hardware synapses. This is a major feature that all ANNs need to function properly.
In the study, published in Nature Communications, researchers had an ANN that employed a memristor to perform reversible learning of cacophonous incoming data.
Memristors normally control the flow of electric current in circuits. They can record the charge that flows through them and can remember the data even with the power button switched off.
If scientists want to somehow manage the complex and next to impossible task of copying the human brain, they will need not billions but trillions of such artificial synapses. Although current technology can allow us to achieve this, the power source and benchmarks won’t be easy to provide.
Novel bespoke synapse components are necessary in this regard. Memristors are a means of exploring this possibility. They can carry out the tasks required of such functionally complex things as human brains (only partially right now) in conditions of extremely compact volume and very low energy costs.
In order to construct an artifical brain in the future will require memristors. They are indeed the royal road to true artificial intelligence in the times to come. The current metal-oxide memristor was able to learn and relearn various tasks and data.
The system used at present for learning and accomplishing tasks is fairly primitive compared to the real wetware stuff that is the human brain. Its marvelous complexity and superior functioning cannot be matched by any machine made by the hands of man.
Any novel technology is a reaching out in the dark and unknown recesses of organized knowledge for something which might work. The circuitry and program, that is the hardware and software, required to run this artificial brain is a monumental task in reality. While things look good on paper, in the real world, they have to be practical and workable to be a success.