***Artificial intelligence***

A few days ago, the 255-year "GO" of human civilization led to the end of the game's dominance of artificial intelligence program Alpha Go. But today's struggle is not a fight for humanity. This is a struggle versus instrument, program versus program, artificial intelligence versus artificial intelligence.
Alfa Go's creator Depmind Technologies recently released alpha go zero of the new version of alpha go. They claim that this version can really shatter the previous version, alpha go zero in 100 matches won 100 matches. This win has taken this device to the highest place in AI Technology. Where alpha-go was taught by utilizing many human knowledge, Alpha Ge Zero itself is self-educated by acquiring knowledge itself.
Computers generally put the information given to people in the artificial neural network, and from there they get the knowledge of complex programs in the algorithmic process on different conditions. After the game of Alfa Go's repeatedly playing with world-famous experts, and the repeated play against themselves, this artificial programmatic knowledge has increased a lot. Where Alpha is the world's top "go" player to GE, shortly after Lee lost another master master like Sedul. Microchip specially made by 48 tensor processing units (TPUs) designed specifically for neural networks.
There are 8 TPUs and a neural network in Alpha Ge Zero, which does not have any idea about "Go", which does not require any kind of observation and it is essentially playing itself against itself, and can quickly understand its own moves and give it And as soon as the master of "Go", the knowledge will increase as much as it plays, which is much more than the previous version's learning system. Many times fast and fruitful.
Where does this alpha go zero knowledge range stop? Or will not stop!