What is artificial intelligence?
Artificial intelligence (AI) is a subfield of computer science that deals with the automation of intelligent behavior. The challenge behind this definition is to clarify what a universally valid definition of intelligence is.
AI can be roughly divided into four sub-areas:
- Human action,
- Human thinking,
- Rational action and
- Rational thinking
The first point includes the well-known “Turing Test”, as a machine perfectly simulates human action here. Alan Turing wanted to test whether a machine (i.e. a computer) has a thinking ability equivalent to that of humans. In the test, a person asked questions that he entered via a keyboard without hearing or seeing the candidates. In this way, he built up communication with a real person and a computer. The test was considered passed if it was impossible to distinguish which of the two counterparts was the computer and which was the human. Since it is very complex to build such a machine, no machine has yet managed to pass this test.
Point two, human thinking, aims to develop machines that can simulate the thinking process of a human. So it’s not just about getting the solution, but about the way the solution was found. Modern image recognition programs, meanwhile, fall into the area of rational action. Aristotle’s syllogisms belong to the fourth area, rational thinking. In the logical conclusion, a true conclusion is ideally drawn from two true premises. An example of this is:
Premise 1: Socrates is a man.
Premise 2: Men are mortal.
Conclusion: Socrates is mortal.
Part of the philosophical debate of AI deals with the differences between a strong and a weak AI. The strong AI is what we know from science fiction. A machine that can solve problems of any kind and answer any question we ask it. However, it is still pure fantasy and will remain so for a long time. The weak AI is what we deal with in everyday life, i.e. complex algorithms that have previously learned their solutions independently.
What is the difference between a normal program and an AI?
A normal program is rule-based and consists of a set of arbitrarily complex instructions:
If this, then that
If the user presses the button, open this application/send an email to server X/…
With artificial intelligence, not every step is explicitly specified, but an algorithm is developed that is able to learn these steps independently.
An example is image recognition, as used in social media, for example. The AI is able to recognize people in pictures, regardless of whether it is a picture at night, in the summer on the beach or in the middle of winter covered with a hat. A set of instructions is written with which it can be learned from many pictures how a person looks. Thus, the AI still does not know all the pictures of a person, but it is able to set up a rule to identify this person even on new pictures. It is important to note that the AI can only learn what it was told to learn beforehand. It is therefore more powerful than rule-based programs, as it can react to unknown situations (e.g. new images) and learns from experience.
“Systems that use artificial intelligence are all around us, from facial recognition to voice assistants to navigation systems.”
Rhea Funke
What are the areas of application for artificial intelligence?
AI is now used almost everywhere. It is able to extract information and patterns from data that we as humans cannot recognize. It should be noted that an AI has no mind. So if it has only received insufficient data to learn or has been poorly programmed, it also draws wrong conclusions and does not notice them.
Examples from everyday life are facial recognition in social media or speech recognition such as Siri or Alexa. But AI is not only used in the online world. Vacuum cleaner robots, navigation systems or autonomous driving are also based on AI. In the geological field, it can help to warn of building collapse in the event of earthquakes and it is also used in the medical field, such as medical imaging. There, the AI supports the radiologists in the interpretation of the image data.
Conclusion
The use of artificial intelligence is already established in the professional environment and has now become an important part of our everyday lives. Systems that use artificial intelligence are all around us, from facial recognition to voice assistants to navigation systems. They are able to react to new situations and learn from the collected experiences, which distinguishes them from normal, rule-based programs. Due to this decisive advantage, more and more application possibilities for artificial intelligence are opening up across all industries.
