To be Uploaded
In today’s world, Artificial Intelligence is one of the leading areas of interest to research scientists.
Research is being conducted with the objective develop software and hardware with cognitive abilities similar to those of human brain. Researchers and scientists have developed extremely tiny gadgets that can simulate the connections between neurons in the human brain better than any that has been invented before.
Consequently the human race can consider the possibility of machines which are like the human brain in that they have human-like capabilities to tackle and solve problems. With this kind of progress in inventing and creating human-like machines, very soon it will be possible to have a higher quality and more efficient and capable form of robots,self-driving vehicles, data mining and analyzing equipment, medical diagnosis and treatment machines, stock-trading analysis and other smart human-interactive systems and machines.
Understanding the working of the human brain
How does the human brain work? How does it have such fantastic and such gigantic computing skills? Let us try and understand this:
The human brain derives this extraordinary power to calculate and compute from the innumerable connections that string its cells together. Earlier research has indicated that the human brain roughly has 100 billion neurons, and that there are around 1 quadrillion (1 million billion) connections that hold these cells together. These connections in the brain are called synapses. At each of these synapses, in general, a neuron fires around 10 times every second. This means that the human brain can execute 10 quadrillion operations per second. Amazing isn’t it?
55 quadrillion computations per second
But if we are to look at Tianhe-2 in China, which is the world’s fastest supercomputer, – this can perform a humongous 55 quadrillion computations per second (this is as per the TOP500 project, which ranks the 500 most powerful computers in the world).
Research done earlier also showed that the human brain consumes only about 20 watts of power – just about enough for a dim light bulb. In contrast Tianhe-2 consumes around 17.8 megawatts of power, which can run about 900,000 such light bulbs, again according to TOP500. The human brain’s capability pales in comparison to this.
Artificial synapses are similar to a transistor
Today, scientists are on a quest to develop computers that can fully imitate the efficiency and power of the human brain. For this they need to invent artificial synapses that can replicate the behavior of the biological ones. So far artificial synapses have consumed more power than biological synapses which consume 10 femtojoules every time a neuron fires. Tae-Woo Lee, a materials scientist at the Pohang University of Science and Technology in Korea and his colleagues have developed artificial synapses that require only about 1.23 femtojoules – the lowest so far.
These new artificial synapses are similar to a transistor where a switching on and off, imitates the firing of a synapse. By creating very small synaptic transistors it is possible to lower the amount of energy they consume. These tiny machines are made with one kind of organic material wrapped around another. These materials help the artificial synapses catch or release electrically charged ions, imitating the actions of biological synapses. Lee and his team are now experimenting with other material to further reduce synaptic transistor energy consumption. They are hopeful that artificial brains will one day overtake human brains in all aspects.