https://en.chessbase.com/post/challenging-studies-automatically
Chesthetica, an automatic chess problem composer, has been capable of composing all sorts of puzzles using the Digital Synaptic Neural Substrate (DSNS) computational creativity approach for some years now. I feel compelled to remind readers that this new approach has nothing to do with machine learning or artificial neural networks. Creativity is typically not something that can be ‘taught’ or ‘learned’. It also has nothing at all to do with ‘deep learning’.