In February 1996, Deep Blue to become the first computer capable of defeating the world chess champion with slow pace, Especially for Garry Kasparov, one of the legends of this sport. Since then, the scientific community has learned artificial intelligence so it can be imposed on people in different areas.
Among these projects is AlphaStar. Artificial Intelligence, developed by DeepMind (Google), which has been dedicated to professional teams of various video games for several years to defeat them. Having shown his skills in games like I'm going, shogi or classic chess, This computer faces a team of classic Starcraft II.
The first machine capable of winning the Starcraft II professionals
As announced by the company through a statement posted on its website, AlphaStar played a tournament last December against Starcraft II's professional players. Genova "Mana" Kominz and Dario "TLO" Wynn. All in order to show off their skills and become the first machine capable of defeating a professional player of this video game.
And he got it for the first time. Although still under development, the machine approved a strong 5: 0 to "MaNa", considered to be one of the best players in the world in this video game, thanks to the so-called neural network. Then came TLO, who managed to beat the machine once, but had nothing to do with his digital rival. Thus, AlphaStar defeated the duo with an aggregate of 10-1.
Special Tournament Terms
As AlphaStar trains only against the race of the Protoss, professionals had to play with this genre to adapt to their style of the game during the development of the tournament. While MaNa seemed more prepared for his confrontation, as he played with this race as usual, TLO was playing in a disadvantaged position, as he usually competes with the Zerg race.
Thanks to their experience against these species, intelligence is imposed on professionals with obvious ease. A performance that surprised players who highlighted AlphaStar's very human way of playing: "I was impressed to see AlphaStar make advanced moves and different strategies in all games using a very human way of play I did not expect. play depends on getting your opponent to make mistakes and take advantage of human reactions so this gives me a new perspective on the game. "