
If we have a smart system that can distinguish a bona fide customer from a bad one, then we can use it, for example, in a bank, to decide whether to give a person a loan or not. Or, if we have a system that can quickly estimate the likelihood of an insurance accident — for example, that a person will be in an accident this year, given their previous driving history — we can use that in an insurance company.
Businesses benefit from the use of artificial intelligence for such point problems, because AI algorithms can take into account many factors that a human cannot perceive at the same time. That is, if we have an event that can be affected by 1000 different factors, then it will be quite difficult for a person to describe an exact formula that would take all these 1000 factors into account. While we can give the machine a large amount of data with previously known answers, and it will independently derive the necessary rules. Accordingly, if you have smarter algorithms, you are better than the competition. If you are better than the competition, you earn more, you have more market share.

Let's recall the example of 1000 factors that I gave earlier. Let's say we are a hospital and we need to decide whether a patient can be given a certain antibiotic given his medical history. If there is a lot of information in the medical history, the human doctor may not know or notice some important details - for example, that the person once had a disease for which it is strictly forbidden to use this antibiotic. At the same time, the machine, having learned from a large volume of stories, can record the connection between this disease and the side effect of the treatment. So, by using narrow artificial intelligence, it is possible to avoid errors related to the human factor and use a smart system as an advisor.




