.
Artificial intelligence and generalizing things
The machine imagination
The virtual simulation that can create the imagination or abstract thinking for artificial intelligence can be already in here. That ability can give artificial intelligence an ultimate ability which makes it closer to humans than ever before.
The computer might have the ability to make simulated games backward, and that ability can call as "machine imagination". In the generation process of the tactics, the computer uses the data, which collected from the opponent's playing style. And then the computer would play the virtual match against the opponent.
In that case, the computer would play things like chess backward, that the user or opponent of the computer would not see, that the system is creating tactics. But when we are saying that only one man can win the chess program, we can ask how generalize this thing is? How probable is that a human player can win the computer if the only winner is the world champion?
Chess and artificial intelligence
Normally we are saying that the human is still winning artificial intelligence in chess. Or modern artificial intelligence with machine learning mode can always win humans, so that thing is not relevant anymore. But then we must ask how to generalize the proposition that chess is the game, what shows that humanity is better than a machine? Have you ever play chess against a chess program?
It's not very easy, because the computer can calculate the movements of buttons far away in the future, and that thing makes winning even the classic chess program, which has not learned mode very difficult if you are an ordinary person, who would not play chess all the time. And when the power of the computers increased, the chess programs turned more complicated, because the computer can calculate more movements to the future than weak and slow machines.
The difference between the power of the processors was easy to see when the same chess program run by using a base PC, which had 640 kb memory and 086 processor. In that case, many people could win that program, but when the program moved to the 386-based computer that computer could calculate so many movements for the buttons, that most of the people would lose. In the cases where the human met chess program the only man who won the IBM chess program was Garry Gasparov the ultimate world champion of chess.
But is the thing, that only one man can win the artificial intelligence in the chess generalized, that the entire population can win the artificial intelligence? And then Gasparov loses to artificial intelligence. The reason was that machine intelligence had the mode where it collected the data of the game style of Gasparov.
Then the computer played the simulated chess matches against the master, and then it was able to create tactics, which made Gasparov lose. That was open the road to the creation of machine learning. If the computer would use the virtual simulation of the game for creating better tactics we can say that computer has imagination.
How to teach a computer in chess? The key element is to find many types of opponents for artificial intelligence. By using this method, it can get as diverse opponents as possible for creating as diverse as possible tactics.
The computer might have the ability to make simulated games backward, and that ability can call as "machine imagination". In the generation process of the tactics, the computer uses the data, what is collected from the opponent's playing style, and then the computer would play the virtual match against the opponent. For creating perfect tactics the chess computer would need multiple opponents, that it can create diverse versions of the tactics, and that's why the chess-pages are the ultimate tool for collecting data for artificial intelligence, what mission is to win human in that game.
In that case, the computer would play things like chess backward, that the user or opponent of the computer would not see, that the system is creating tactics. But when we are saying that only one man can win the chess program, we can ask how generalize this thing is? How probable is that a human player can win the computer if the only winner is the world champion?
Image: https://en.wikipedia.org/wiki/Outline_of_chess
Comments
Post a Comment