Scientists have found the key to recreating the human brain?
If you want to build a computer on what that image and likeness, then in that case you have to copy it exactly. If you have a desire to build computer based on human brain, you need at least to copy the process of the brain. All you need to understand and realize to the smallest detail. If you believe IT news, it is probably the scientific world is very close to the successful implementation of idea. Scientists have already created real change neural connections and on the prototype showed how the electronic signal travels in the brain. Already, bio-engineers understand how the human brain stores tons of information and how it is quickly of the common data Bank pulls out and uses. It is in fact normal electrical signals in our brain that allow us to see, process and remember.
Scientists have managed to repeat all three stages of signal processing in the brain. Earlier attempts continued, but not processed, the signal came for the end point with the same speed as the processed. Parallel computing using a new approach can be done in unlimited amounts, so that processing large arrays of simple data can be done very quickly, so quickly that to represent the process itself is very and very difficult.
The laboratory developed an analogue of the synapse, simply and tastefully. Conventional double-layer battery and heaps of wires to the sensors in the salt solution. The idea itself, the device operates in key mode (the key is assembled by analogy with the transistor and to him are three wires), but the end is needed in order to control the signal level. The voltage and current is maintained at the same level as the human brain, a comprehensive recap of the chemistry of the brain.
The purpose of this experiment is to create a neural network which would have fulfilled a large part of the processes that take place in the brain. The first step in achieving this goal is passed.
Created while only one synapse, but his example was the real opportunity of the entire system. If you collect at these synapses, we obtain a neural network which can recognize even the text is hand-written by the man, the precision error will not exceed 97%. Everything seems quite simple, but there is one problem and it lies in the minimization of the artificial synapse, the more dimensions, the more energy is needed for its stable operation. Yet if collected on artificial synapses bio computer, it would consume energy in 10 thousand times longer than processing the same task would take a human brain. If scientists still minimise its neural network and bring it in size to real, energy will be spent less commensurate, and opportunities to use will be much more.
<<<< Back >>>>