The term 'artificial intelligence' (AI) is not an accurate technical description, but rather a social narrative. The current AI debate is primarily about large language models such as ChatGPT.

 AI is technology, not magic. The output of large language models is essentially based on probability calculations based on huge amounts of data.

The current development of AI seems breathtakingly fast. However, this technology has actually been researched and developed for a very long time. Much of it is only now becoming accessible to a wider public.

Large language models are based on human labor. In particular, they incorporate content designed by humans, programming developed by humans and classification and filtering work carried out by humans.

What seems human in AI is often actually human. Example: ChatGPT often responds very compassionately when sad events are entered. However, it was actually people with compassion who programmed ChatGPT accordingly.

Large language models reproduce the past. If you don't like many things from the past (e.g. the prevailing stereotypes and prejudices), you have to become aware of this reproduction, recognize the individual characteristics and make targeted changes.

Who benefits from AI language models and who suffers from them is very unevenly distributed. The Matthew principle very often applies: to the one who has, will be given (and vice versa!).

 Education is aimed at smarter people, not smarter machines. The exciting question is: (How) can machines help us to become smarter and act smarter or how can we become smarter and act smarter in interaction with machines?

Poor education remains poor education - whether with AI technology or without. The goal must not only be 'learning with AI', but above all 'learning in and for an AI-influenced world'. This is not primarily a technical question, but a question of learning culture.

Further processing of content often works much better with large language models than the generation of new content.

AI technology is not neutral. It is part of social negotiation processes. That is why it is never just a tool, but always an object of reflection and design.

Dealing with AI technology requires learning across society as a whole. In many respects, AI development is a contradictory process. We can make progress together if we enable more 'both-and' instead of 'either-or'.