Artificial intelligence is a complex concept for which there is no unambiguous definition. The concept of artificial intelligence (abbreviated as "AI" for convenience) is used by specialists in various fields: writers, journalists, business and science; and different specialists put their own meaning into this concept.
In the broadest sense, artificial intelligence is the ability of a computer to solve the same intellectual tasks that a human is capable of solving.
This concept can be specified at different levels:
Today, there is only highly specialized artificial intelligence: a machine capable of solving one given intellectual task (for example, face recognition, playing chess or machine translation). At the same time, it is unclear whether it is possible in principle to create a general artificial intelligence using existing technologies, that is, a machine that can solve various complex intellectual tasks.
Let's take a personal assistant: general AI implies that the machine will be able to replace the assistant entirely (it will perform various kinds of tasks, schedule, answer calls, and so on, and will also be able to independently learn new tasks). The technologies that exist today are much simpler: the machine can record text by voice input, offer short answers to incoming letters of the form "Accepted for work. Sincerely, (name)", or set reminders based on a text or speech description, with a separate algorithm responsible for each listed function.
On the other hand, the creation of algorithms for performing even these tasks, which seem simple in contrast to the functionality of a full-fledged personal assistant, historically took a lot of time.
Artificial intelligence includes many areas of mathematics and information technology, as well as biology, physics and other sciences. We have described only the most famous and significant events and approaches, of course, there were much more ideas for creating artificial intelligence. For example, there is an approach based on evolutionary algorithms: it consists in trying to simulate "evolution" using random "mutations" of the program; today such approaches are used in conjunction with modern artificial intelligence systems, such as neural networks.
In addition to understanding artificial intelligence as the ability of a computer to solve intellectual tasks like a human, there is an understanding of AI as creating a computer that simulates the human brain. However, science does not yet fully understand how the brain works, so there are no ways to artificially repeat it either.
To date, artificial intelligence and large data processing technologies are actively used in business, the next block of the course is devoted to specific examples. At the same time, there are several related areas responsible for the development of these technologies.
A class of artificial intelligence methods, the characteristic feature of which is not the direct solution of a problem, but learning in the process of applying solutions to many similar problems.
Sometimes called "deep learning" (from the English Deep learning). A subdomain of machine learning where neural networks are used as algorithms.
It is a concept of combining statistics, data analysis, machine learning and related techniques to understand and analyze real-world phenomena.
A broad concept that means extracting knowledge from data.
This is a set of approaches and methods developed for analyzing huge amounts of data.
It is worth noting that not all data work relates to artificial intelligence (for example, an analyst making a conclusion from graphs does not relate to artificial intelligence), and not all artificial intelligence algorithms are developed using data (for example, expert systems mentioned above). Nevertheless, many modern intelligent systems are based precisely on data learning: machine translation, image and speech recognition, forecasting customer behavior, etc. During data training, the algorithm "studies" a large number of real cases (for example, customer behavior or text translations) and thus makes qualitative predictions for new cases.
To summarize: historically, the concept of artificial intelligence has been transformed many times, and today artificial intelligence technologies are largely based on learning from large amounts of data, but in addition they also include complex calculations, expert systems and other algorithms.