The term artificial intelligence (AI) seems to be ubiquitous these days. It is being used or is supposed to be used, in almost all areas to automate processes, perform difficult tasks or to help us make decisions. In my talk, which was presented at Code.Talks 2022, I invite you to take a look behind the hype around AI. From my practical experiences as a data scientist, I will show you that many relevant use-cases can be solved without AI, just by applying some old-school mathematical modelling. You will see the many advantages of mathematical modelling like interpretability, data-efficiency, robustness, etc. over AI-based approaches. The main take-away is that you should always focus on understanding the problem itself before you choose a method - be it a neural network or linear regression - to solve it.