The simulation algorithm of the human brain already exists, but no computer can do it

<pre>The simulation algorithm of the human brain already exists, but no computer can do it

An international group of researchers from Japan, Germany, Norway and Sweden performed one of the cherished dreams of an enlightened humanity. They presented a technical paper that details the algorithm for modeling the work of connections between neurons and synapses in the human brain. In simplified form, this is the key to recreating the brain's work on the computer.

The neuron is the basic unit of the nervous system, the cell with the information signal transmission mechanism. The synapse is a node of neurons, the place of their connection for the exchange of information. The combination of neurons and synapses forms a structure that generates the functionality of this part of the brain: memory, data processing, decision making and much more. Modeling all such possible connections and analyzing the result is the ultimate goal of the new algorithm.

One problem, this requires extra computing power. It is generally accepted that there are 1 billion compounds in the human brain. Even if we describe each neuron in just one bit, the whole brain simulation will require a memory capacity that is 100 times larger than the volumes used in modern supercomputers. In practice, all this kind of goodness needs to be put together to work together, and this is additional capacity.

As a result, the implementation of the developed algorithm is not yet possible, but the merit of the team of researchers is that their brain model is made “extremely scalable” and in open source code. And with the growth of computing power, it will be easier to achieve a simulation of the brain's work, until one day we reach 100%. And then humanity will enter a new phase of its development.

           js.src = “&version=v2.8”;    'script', 'facebook-jssdk'));

Source link