When we try to solve problems, we usually use tools such as search engines, read books and references, ask colleagues or experts, sometimes even people who explicitly have no knowledge of the problem domain. This can be understood as a creative process, because when an answer is not immediately apparent, information must be processed, familiar concepts reinterpreted, structures rethought, and ways of thinking overcome. Not only are new answers learned, but new neural connections are also formed in this process. And in this process, information is also processed that is not necessarily, and sometimes not at all, related to the actual problem. Although this information does not lead to the solution of the specific problem, it is still absorbed through processing, broadening horizons and allowing past and future problems to be viewed from new perspectives. This is where the special value of traditional work with search engines, forums, and answer sites lies, where people look at and answer a question in different ways.
In short, collateral knowledge can be understood as knowledge that was unintentionally acquired in an unrelated process.
AI usually provides us with a single answer with exaggerated and misleading self-confidence. Studies have also shown that outsourcing the thinking involved in problem-solving to AI permanently reduces the user's brain activity, and thus also their ability to solve future problems more effectively. Not only do we no longer activate our brains, we are permanently destroying the thinking ability of our brains, which we need to solve future problems. Furthermore, the answer provided by AI gives us a direct path to the solution without presenting different perspectives. There is no differentiation of the question in terms of the solution to the problem, sustainability, efficiency or effectiveness, moral and ethical perspectives, humanistic values, and potential dangers through abuse or misuse. In the end, there was no journey of the problem solver which required them to look left and right. There is no reason to investigate further, if the first solution the AI presented, works. No further thought is spent analyzing the given solution. It is copied, pasted, tested, and will be forgotten. So in the end, using AI to solve problems reduces our own capabilities to solve problems, as well as it prevents learning about things, that are not related to your problem, which you require for reflection and creativity.
This text was authored by Oliver Eglseder Dec. 2025