After the recording of the first demo with my band Mundo Kumo where we create music inspired on per reviewed papers about climate change, it was time to create some of the videos that we will be using on our live show. My initial idea was to use computer generated art combined with real climate data. I had experience with coding and complex artificial intelligence algorithms however not much experience using artificial intelligence (AI) to generate art.
The code used here was developed using the python language and it can be downloaded freely on my github repository called AlgorithmicArt. Although this was the initial idea about the videos, the idea changed a lot after and we are not using all the videos explained here. However they are still a really good starting point.
When doing a little bit of research on the topic i found that Cellular automata was probably one of the easiest way to start. The concept of cellular automata was created by the mathematician Stephen Wolfram and some of the rules are very similar to the majority vote model used widely on statistical mechanics.
The first intuitive step is the use of the one-dimensional cellular automata (the easiest to implement and understand) . Thus picking a cellular automaton rule, it is possible to create a picture similarly to the pictures found on the internet.
To make things a little bit more interesting, a video can be made showing the evolution of the cellular automata grouping sequential gifs. This can be done using different alternatives. For example using ubuntu and the imagemagick package:
convert -delay 20 -loop 0 *.jpg myimage.gif
The final video is a thing like this:
The second step is to add the climate information in the code and combine with the cellular automata. I used global temperature anomalies from NOAA. I maped the colors based on the range of temperature anomalies. Thus, the colors (global temperature anomalies) and the evolution of the cellular automata are connected:
Next post I will talk about two-dimensional cellular automata.