Dimension reduction is a complicated topic with a vast zoo of diverse techniques for different specialised problems. This talk will seek to cut through the technical detail and focus on the core intuitions that lie behind dimension reduction. From this point of view we'll see that there are only really two core ideas you need to know to understand dimension reduction. With those ideas in hand we'll look at how popular dimension reduction techniques such as PCA, NMF, LDA, t-SNE and UMAP can all be seen as specialisations of one or the other of those two core techniques.