Artificial intelligence is something we have tried to manage for decades, and we are getting better and better every day. Overall you can describe artificial intelligence as we are trying to combine the human brain with a computer, we want our computers to be able to think by them self and think like humans.
History of Artificial Intelligence
In the late 1800 many scientists had thoughts about artificial intelligence and the invention of artificial intelligence was at its beginning. The scientist claimed that they had an answer to what artificial intelligence was. They imagined a programmed machine that automatically after a specific procedure could perform instructions from logical thinking.
In the early 1900 the artificial intelligence was still just thoughts the scientist came with, one of the most interesting thoughts was the one Alan Turing came up with in 1950, were he claimed that he had a way to test a computer whether it was intelligent or not. The principles in the test were really simple, the computer and a human should have a chat conversation with the judge of the test, and if he could not tell which chat conversation was with the computer, then the computer had to be intelligent. And that is how the Turing-test occurred and still is in use to this day, but it was still just a thought, we have to go a few years into the future before anyone tried to create artificial intelligence.
After world war 2 the world’s first programmable digital computers were invented. The invention of these digital computers made it possible for the first time to search for artificial intelligence. In 1956 a group of computer engineers decided they would design a new generation of intelligent computers, the supercomputers. The group of engineers had a lot of confidence in themselves that they could easily build a super intelligent computer, but the result was not what they had expected, instead of making a supercomputer they made computer programs that could solve mathematical problems. Because of the bad results and the development of artificial intelligence did not go as fast as expected, the program of developing artificial intelligence did not receive a lot of funds in the following decades.
In the beginning of year 2000 the artificial intelligence started on an international scale, but they did not research for the human artificial intelligence. Instead they split it into a lot smaller scale and started to work on machines that were good in specific areas. In these specific areas the machines were way better than the humans. They were able to build a chess computer, that was able to beat the leading human chess player, a car that could drive by itself on a desert road without crashing. In 2011 the scientists had a breakthrough when they made a machine that was better and faster in general knowledge than the world champions in jeopardy.
Strong and soft AI defined
In the big picture of artificial intelligence there are two different types of artificial intelligence: The strong and the soft types of artificial intelligence.
What exactly is strong artificial intelligence? In the research of strong artificial intelligence there are two different divisions: The human aspects and the logical aspects and these two divisions are also split in two divisions, how the machines think and how they behave.
Artificial intelligence: 1.Human / 2.Logic
1.Think Think like a human / Think logically
2.Behave Behave like a human / Behaves logically
1.1 ) The machines have to think like humans before we can say that they have some sort of artificial intelligence. The big problem with this is that is that the humans actually do not know enough about how the human brain exactly works. Right now they do not know enough to create a bridge between the human brain and the computer.
1.2 ) The machines have to think logically, that means the computers have to control their thinking to make the right decision in the given situation. Based on what their senses tell them is the right answer.
2.1) When we are talking about the machines having to behave like humans, we can take a look back on the Turing-test if they can pass the test they can also behave like a human. In the human behavior there are a few criteria’s the machines have to meet, they need to have a good knowledge of the language, so they can write correctly, they also need a good memory so they can remember what they have been told, the ability to learn is also important, so they have the opportunity to adapt to new situations.
2.2) Logical behavior is hard to describe, the machines need to find out what the right thing is in the given situation. As an example could it be about a dilemma between right or wrong, is it okay to kill if it is the only way to not get yourself killed. These decisions are also difficult for a human to make so it is really hard to create a machine that is able to make that choice.
Today the strong artificial intelligence is still an area we do not know a lot about, it is still something we are working on in a small stage. It is a long-term process to come from where we are right now to accurately have built a machine that can control the aspects in the strong artificial intelligence. Every year there are contests for the best strong artificial intelligence machine where they compete in which machine that can beat the Turing-test and so far there is not a machine that has beaten the test.
Soft artificial intelligence is defined as not sentient, which means that the machine is programmed to do a specific thing. In that specific area the machine is very good, sometimes they can be a good replacement for a human because they have the same working abilities as a human, but are able to maintain high-performance for a lot longer period than a human.