AI and Machine Learning series
How can German companies still become digital masters? How can this be initiated and driven forward using the means of corporate communications? Why do we need less rather than more information? How can decision-makers and experts develop pragmatic approaches that are actually feasible? - These are the questions of my focus on digitalization this autumn here at PR-Doktor. In dealing with the topic, I realized how much catching up I still had to do in terms of knowledge. That's why I asked a machine learning expert to answer my most burning questions. The huge amount of answers, information and examples has resulted in this three-part series. - Kerstin Hoffmann
Part 1: What decision-makers need to know about artificial intelligence and machine learning (this article)
Part 2: Danger in Imminent Danger? The Future with AI and Machine Learning
Part 3: AI and Machine Learning in Marketing and Corporate Communications (coming soon)
The term artificial intelligence (AI) was first coined by John McCarthy in 1956. The term is roughly described as a machine that is capable of performing tasks that are characteristic of human intelligence, such as planning, understanding language, recognizing objects and sounds, learning, and problem solving.
There are essentially two areas of AI, generalized and bc data special (called general and narrow in English). The generalized area basically does almost everything that human intelligence can do. The special area is where highly specialized things like object recognition come into play, for example when the machine really only recognizes images. What we currently refer to as machine learning is mostly classified in the area of special AI.
Reduced to the basics, one could also say: Machine learning is the foundation for the creation of AI.
But that sounds familiar at first. IT has been providing solutions for all of this for years. What is the specific difference to previous information technologies and what is the enormous potential of machine learning?
The difference is as simple as it is significant. While familiar information technology usually describes the problems of our everyday lives approximately by describing more or less complex sets of rules, mathematics and similar procedural models, machine learning only needs the data from the world to be described. It learns the rules on its own! The machine independently creates a model of the world to be described, based only on the information generated by this world.
In addition, the machine is able to improve itself over time. This usually happens through targeted feedback from users or systems.