The current "AI" is much more A(rtificial) than I(ntelligent). You can look at it as pattern matching functions based mainly on linear algebra - so how should that threaten humanity? These tools have no internal goals or purpose, no ambitions whatsoever. Such, they can not have a conflict of interest with us humans. And even AGI, which is not yet existing but certainy will one day, will be just a tool which can reason in a rational manner. Again, it will not have any internal ambitions. The dangers always come from humans missusing the tools. (Look at kitchen knifes: they are very helpful for houswifes and cooks, but more people are killed or injured with them than with anything else.) In contrary: AGI will allow to identify and control or even stop misuse of it - certainly in a better way than any technology ever before was able to.
artificial intelligence is not a self-developing system. The differences in thinking between the human brain and the computer are too great.
Artificial intelligence is an attempt to imitate human thinking, and we do not know exactly how it happens. Therefore, today artificial intelligence is a set of technologies in applied fields: pattern recognition, automatic translation and a number of other tasks.
The systems currently known as AI are able to solve one task, such as play chess, play Go, object detection on images, natural language processing, among others. These systems are also known as weak AI or narrow AI. However these systems are not able to self-evolve.
I think what you are talking about is what it is known as General Artificial Intelligence (GAI) or strong AI. GAI systems will be able to do all the tasks a human can do at least at the same level as a human. In theory, GAI will be able of self-evolving, and it will do it very quickly. This will be known as Intelligence Explosion. This will bring a lot of changes in laws, weapons and the job market.
I recommend you to read the book from Professor Max Tegmark, Life 3.0: Being Human in the Age of Artificial Intelligence. He explores the topics you are talking about very deeply.
The current "AI" is much more A(rtificial) than I(ntelligent). You can look at it as pattern matching functions based mainly on linear algebra - so how should that threaten humanity? These tools have no internal goals or purpose, no ambitions whatsoever. Such, they can not have a conflict of interest with us humans. And even AGI, which is not yet existing but certainy will one day, will be just a tool which can reason in a rational manner. Again, it will not have any internal ambitions. The dangers always come from humans missusing the tools. (Look at kitchen knifes: they are very helpful for houswifes and cooks, but more people are killed or injured with them than with anything else.) In contrary: AGI will allow to identify and control or even stop misuse of it - certainly in a better way than any technology ever before was able to.
I believe humans are superior creatures, they have the final say when it comes to technology and artificial science. The day when AI systems over control us won’t arrive unless humans themselves plan to witness it.:)
The term artificial intelligence comes from Alan Turing. In 1950, he published the famous article "Can a Machine Think?". He also proposed the criterion of artificial intelligence, today known as the Turing test.
Turing formulated his test like this: you communicate using the keyboard with the other person. You can ask any questions. If within 10 minutes, you do not understand who is talking to you: a computer or a person, then the computer passed the test and the computer can think like a person.
Alan Turing predicted that during a five-minute test, a computer with a memory of about 126 MB could deceive a person in 30% of cases. Today computers have moved far beyond the borders. In the Internet there are many chat bots.
Over time, it came to be understood that the Turing test and artificial intelligence are slightly different things. In today's understanding, the Turing test makes it possible to evaluate only the imitational capabilities of a program — that is, how well it pretends to be human.
The term “artificial intelligence” dates back to 1955, when John McCarthy offered it at a conference at Dartmouth University.
Against ardent supporters of AI, his famous “Chinese room” argument was put forward by Searle (Searle, 1984). By analogy with a computer, he suggested a room in which workers who do not know Chinese translate texts from Chinese into English slavishly following a set of rules. Searle, making an analogy, said that no AI program can understand its subject matter.