A part of the magic of artificial intelligence is how they understand the world and how things relate to each other. In this talk we look into how different advances has been made in machine learning, from seeing written numbers in the 90s via the then shocking take-over of ai in image recognition and to the language models of today. Through this journey we will learn how neural networks see the world, their understanding of how pieces of the world relate to each other and how they produce astonishing results in several different applications. Let's take a trip through the embeddings and in the semantic space!