Artificial intelligence use in every aspect of our lives these days. In the future, the role of ai in our lives will expand and the future will become a world in which we cannot live without ai. However, artificial intelligence is just a computer program that realizes what human intelligence can do. The computer can recognize only ‘on’ signals with current flowing and ‘off’ signals with no current flowing, and expresses ON as 1 and OFF as ‘0’. But how can a computer that only knows 0’s and 1’s become artificial intelligence?
What is artificial intelligence?
A signal representing either 0 or 1 in a computer is called a bit and a bit smallest unit of information processed by a computer. Since a bit is either 0 or 1, there is very little that can be represented by a single bit. So, 8 bits groups and processed at once, which is called a byte. Using these bytes, the computer can display information such as numbers and characters. A byte can represent 256 pieces of information since it can multiply two signals of a bit (0 and 1) 8 times. So, one character expresses by one byte for an alphabet, two characters by one byte for a number, and one character by one byte for a special character.
You can create various languages that make programs with the letters and numbers expressed in bytes and write commands in this programming language one by one to create a program. When this program replaces what human intelligence can do, we call it an artificial intelligence program.
The top-down approach and expert system
Over the first 30 years of artificial intelligence research, scientists have been trying to create artificial intelligence in a top-down approach. A top-down approach is a method in which the whole (top) determines the parts (bottom). The top-down approach in the development of artificial intelligence method in which the rules and information related to intelligence stores in the computer, and the computer makes decisions by itself by comparing it with information received from the outside.
This method is an attempt to mimic human intelligence, regardless of how the human brain processes information. Eventually, scientists found that there were limitations to creating computer programs that could replace human intelligence in this way. Instead, they realized that the more knowledge entered into the program, the greater the program’s ability to solve problems. So, from the late 1970s, scientists began to develop expert systems that put a large amount of knowledge in a specific field.
An expert system is a program that puts the knowledge or experience of an expert in a specific field into a computer. So that the computer can make judgments or reasoning like an expert. This program can answer questions about car breakdown symptoms.
Also read: Google changes backup solution in Android
Artificial neural networks and a bottom-up approach
Information provides by our sensory organs, such as our eyes, ears, and touch. Converts into electrical signals and transmitted to the brain. The signals sent from the eyes to the brain cause the same signal patterns among the brain cells.
When we see an object frequently, the same signal pattern repeats between brain cells. AI scientists have created models called artificial neural networks by mimicking the learning process of the human brain. AI creates through the learning of so many artificial neurons, this model calls a bottom-up.
Computers only operate according to the rules of programs created by programmers, and cannot change rules or come up with new rules on their own.
Computer programs using artificial neural networks learn from mistakes and correct their own errors. The learning process of the artificial neural network is as follows.
There is a high probability of giving an incorrect answer, but errors reduce through trial and error. In the end, it gives the answer 3 and remembers the path that gave this answer. The number 3 written in other handwriting is also learned in this way, and eventually, the AI learns that the number 3 is the first handwriting it sees.