Artificial intelligence defined in a nutshell

Big Data - Internet
Big Data – Internet

With all the euphoria and hype, the term artificial intelligence belongs (Of) first explained.

To the chagrin of science, there is no exact definition, which is why many things and technologies creep in, which have nothing to do with AI.

It hurts in the head, which journalists sometimes advertise as artificial intelligence. Many a pioneer would turn in their grave.

Artificial intelligence is after the statues of the DFKI (German Research Center for Artificial Intelligence) Teaching human behavior to computers.

A little more technical can also be said:

Perceive -> Understand -> Act -> Learn

The skills described above represent a gross simplification.

If you say now, that even previous computers and mainframes were, according to your perception, productive through input - processing - output, then I agree with you.

The real news about artificial intelligence is learning and understanding.

Thanks to the ability to learn, systems can also be trained and thus achieve better results than with conventional methods, which just stare at, clearly defined and firmly programmed processes (regulate) based.

Weak AI supports people

For a better understanding and limitation, a distinction is made between weak AI and strong AI.

The weak AI is, when it comes to, to intelligently support people in achieving their goals. Human interaction <-> Machine z. B. by an industrial robot, who works with people.

If you talk about technology today, then it usually refers to the weak AI.

The strong AI is still a long way off and is more at home in Hollywood.
In short, the strong AI aims to imitate humans.
There, too, development is unstoppable, But as far as I know, there are currently no application fields for strong AI that are ready for series production.

Pocket calculators don't have artificial intelligence

Inevitably, sooner or later the question arises, whether a calculator also includes artificial intelligence. No!
A calculator does not evaluate any results, does not assign any meaning to results and does not take any independent actions with the results.


Many terms and methods are not included in this article for the sake of simplicity. This would go beyond the simplicity of the article. Das KI-Lexikon is gradually expanded to include the most common terms of artificial intelligence.