They just take it and basically rap it in a rhythm, and then make the visual part of it. Not in three dimensions and two dimensions, but using a very simple AI algorithm. Actually, the result is not bad. It is quite good.
j previous speech k next speech