Latent Solids

Computational Art

An ongoing exploration of the generative potential of Artificial Intelligence (AI) and geometrical shapes.

I’ve been interested in Generative Art for many years now and the possibilities of using Generative Adversarial Networks (GANs) for the creation of visuals and imagery is really exciting. Although we are going through some pretty weird uses of AI in society, I think that’s actually a good push to explore the field in a more creative, less invasive way.

Thinking about the ethics and AI (special thanks to Lia Coleman, Claire Leibowicz, Emily Saltz and Lia Coleman for the great discussion on How to Use AI Art Responsibly”) I’ve started generating my own visual dataset with the 5 classical Platonic Solids and feeding it to a Neural Network. I’m using StyleGAN2-ADA to learn from this data and possibly generate a new perspective on these classical shapes. By exploring the latent space created by this network I’m hoping to understand how GANs make sense of geometrical forms and how can this be interpreted in a more philosophical way.



Concept & Dataset

Sandro Miccoli


Artificial Intelligence, Machine Learning, Deep Learning, Neural Networks