Artificial intelligence: from chess programs to androids

    

.

Artificial intelligence: from chess programs to androids

From chess program to drone swarms

Chess was the perfect game for early artificial intelligence tests because the rules of the game are simple. Also, the gaming area is limited. That limits the number of variables that the computer must handle. Chess buttons can also mark by using RFID-identifier, which are in use in the thief alarm systems in supermarkets. That system makes the computer easier to detect and separate the buttons. The gaming platform can also equip with a detector, which detects where each of the buttons is what makes the computer easier to analyze the game. 

In this version, the computer must not detect the shape of the button that makes the system easier to program, because the recognition of the button happens by using digital code. And that's why programmers often used the chess game as a test platform in primitive artificial intelligence systems. What opened the path to the modern neural network-based systems like independently operating robot groups like drone swarms. 

But modern systems can detect the button by using their shape. That means that the computer can play chess by using a regular chessboard, which has no digital detection system. The system would detect the shape of the button, and find from the database, how the button must move. And then artificial intelligence plays a virtual game on the chessboard in its memory. There is a limited number of movements, and the computer can make all of them in a couple of seconds. This makes that kind of system hard to win because they are playing all possible variations of the game before the first movement is not made In this process the computer makes the series of movements. 

The limited possibility of movements allows the computer can calculate the series of movements before the game has started. 

The idea is that both players have limited opportunities to move their buttons. The computer makes a couple of series about the possible movements of the buttons what it and the opponent can do.  And how the opponent makes the own move or counteracts makes the computer select the best suitable series of the movements. The computer can calculate all possibilities forward and the thing, what the opponent is making describes which of the movement series the artificial intelligence selects. And then it would choose the best one. The modern system would use the camera and image compilation for detecting the right button. 

In chess, artificial intelligence can collect data from a limited area. There are clear rules in that game, which are easy to program in the memory of the computer. The computer is easy to follow the game and the chess buttons are easy to mark for the computer. There are no hidden things and every action on the game happens on the 64 squares, which means the computer doesn't need the drones or other things to find the things. 

Chess is an excellent tool for simulating the situation where the opponent for the computer is human. In this kind of systems, the computer can play chess against human in the limited area, where is strict and quite easy rules. And that thing was really important in the early artificial intelligence systems. The number of variables in chess is limited, and they are very easy to program for the computer. This thing was important in the time, where artificial intelligence run on single stand-alone computers. But then came the internet, which allowed to link the computers together, and that opened the road to independent operating robots. 

Modular programming and artificial intelligence

The modular programming gives new abilities for independently operating robots. The robot can walk on the street and use mobile data transfer tools. It could select between a fast 5G-network and it can also benefit free WLAN systems. When a robot requires instructions it can download the necessary data from the GSM-network, or it can walk in the shopping center, where it will find the open WLAN, what it can use to connect itself to its support network. 

In this text, the independently operating artificial intelligence-based robot means robot, which can operate without human control. The thing is that the robot can connect to a central computer by using a mobile data network, and that allows the robot can make many missions that people cannot even realize. If we are thinking of the "servant android", what can go shop, cleaning the house, or drive the car, we can make that kind of robot by using high-power mobile data and computing. In this type of robotics, the programming is like in module. 

When the robot gets an order for going shop, the command, what is given by speech would activate the module, where is stored the needed actions for going shops. In the shop, the robot would use the scanner program, what it would use to compare the shopping list and products. The thing is that this kind of robot can use the internal internet, but it can also see things from the screen of the mobile telephone. 

From the chess programs is a very long journey to the androids that can walk and operate independently on the streets. The street operating androids can operate independently, and they might look like regular people because their robot skeletons are covered by using synthetic skin, but what means independently operating robots? In this kind of system robot itself is the tool that is remotely connected to the computer system, which might be one single supercomputer.  

But the robots can also create the neural network together. This kind of neural network is used in the independently operating drone swarms, but the shape of the robots is not important. In this kind of system, the robots are creating the independently operating entirety. The idea is that every robot would share its data processing capacity with other members of the group. And that thing makes the drone swarms and other things ultimate tools. 

There is a possibility, that the internal communication between drones in the swarm is disturbing by using plasma fields or high-power ECM-jammers. In those cases, the drones must turn to use optical communication which requires straight visible contact between those drones. The system can use the "outside drone", which would be higher than others, and then the communication between drones happens by using that kind of relay drone. The relay drone would deliver the communication signals of those drones in the case that they require optical communication. The drones can send their laser rays upwards, and the relay drone would resend that data to other drones. 

The training of the robots can happen just by loading the new module for those machines. And this is the reason, why the military is so interesting in this technology. If the robot is working as a security guard of the military base it can check the security passes, but it can just make fake checks. The thing is that the robot can scan the retina of the person or it can scan the fingerprints. And if the wrong person attempts to get in the robot can use the acoustic stun system for arresting those people, who are entering the area without permission. 

Image: https://www.smithsonianmag.com/innovation/google-ai-deepminds-alphazero-games-chess-and-go-180970981/