Generative models such as DALL-E or Generative Adverserial Networks (GANs) are primarily known as accessible, low-cost ways to generate images. Could they also be used to study social phenomena, such as the visual aspects of social processes in cities?
