There are various forms of artificial intelligence (AI) today. It’s a tough question to even call it AI and just call it a software program. There is a trend in software, where something that used to be called “AI” matures and integrates into the technology landscape, no longer being called AI. Programmers in the 1950s might call various programs embedded in our world “artificial intelligence,” for example, the microchip in your car that regulates fuel injection, or the database in the supermarket that stores records of all sales, or the Google search engine.
All AI designs are at least somewhat inspired by the human brain.
But the field that calls itself “Artificial Intelligence” tends to be slightly different from the much larger group of “general software developers.” AI researchers tend to look for forms of software that are more complex, adaptive, capable, or even vaguely human-like. AI workers also tend to be interdisciplinary and well-versed in areas of science and mathematics unfamiliar to the typical programmer, including but not limited to formal statistics, neuroscience, evolutionary psychology, machine learning, and decision theory.
In the field of artificial intelligence, there are two main camps: the Neats and the Scruffies. The division has been around practically since AI was founded as a field in 1956. The Neats are advocates of formal methods like applied statistics. They like their programs to be well organized, demonstrably sound, operating on concrete theories, and freely editable. The scruffy like complicated approaches like adaptive neural networks and consider themselves hackers, putting together anything as long as it seems to work. Both approaches have had impressive successes in the past and there are also hybrids of the two themes.
All AI projects are, at least superficially, inspired by the human brain, since, by definition, AI is about mimicking some aspect of intelligence. AIs have to construct concepts of the things they manipulate or work with, and store those concepts as blocks of data. Sometimes these blocks are dynamic and frequently updated, sometimes static. In general, an AI is concerned with exploring relationships between data to achieve some goal.
Objectives are often assigned based on usefulness. When presented with a goal, an AI system can generate sub-goals and assign utility values to those sub-goals based on their intended contribution to the main goal. The AI continues to search for sub-goals until the main goal is achieved. You are then free to move on to a new (but often similar) main objective. What differs widely between AIs is how all of these dynamics are implemented.
