We frequently use computers to aid us in making decisions. Want to know where’s a good place to eat in your area? Just do a quick search on your phone.
Want to find some information for a class your taking? Another quick search! But what about our devices actually making decisions on their own? What about learning and adapting to different situations?
Computers are certainly capable of making some decisions and predictions as it stands. For instance, if you attempt to carry out a Google search, you will notice that you will be fed with suggestions for the item you are searching for even before you complete typing the search query.
In addition, several email apps these days are capable of sorting important from not so important mail.
Computer systems are capable of amazing things such as defeating the world ‘s best Go player with the aid of artificial intelligence. However, there is still some way to go in terms of carrying out functions such as abstract reasoning.
I suppose the holy grail of artificial intelligence would be to build a system which could emulate the human brain.
It just so happens that researchers from the University of Electronic Science and Technology of China and Nanyang Technological University in Singapore have created a ‘neuron transistor’ which simulates the behavior of a neuron in a human brain.
Such a device could form the foundations of a device which ultimately mimics the functionality of a human brain.
A neuron is a type of cell which can be found in the nervous system of the human body. The human brain contains several billion neurons. Neurons transmit messages and are essentially responsible for controlling the functions of the human body.
Overview of the Neuron Transistor
The researchers created a transistor from a semiconductor material called molybdenum disulfide (MoS2) which is capable of carrying out a counting task similar to re-arranging the beads which make up a two-bead abacus.
A neuron in the brain is capable of receiving signals from other neurons. Based on the information contained in those signals, it will ‘decide’ if to ‘fire’ or not. The neuron transistor must be able to mimic this behavior.
The researchers created a transistor from a semiconductor material called molybdenum disulfide (MoS2). It is capable of carrying out a counting task similar to re-arranging the beads which make up a two-bead abacus.
While there have been other similar devices in the past, their operational speeds have been relatively low. A neuron in the human body fires at a rate of around 5 times per second. Previous neuron transistors have not been able to exceed speeds of 0.05 times per second.
However, the device from the researchers at the University of Electronic Science and Technology of China and Nanyang Technological University can fire at speeds from 0.01 times per second to 15 times per second.
In the future the researchers hope to modify the device so that it is able to carry out more complex tasks.
It should be noted that this work marks only a basic manifestation of neuron function. With improvements it could form the foundations of computer systems capable of making complex decisions and adapting to different situations.