What happens when Artificial Intelligence is the artist?

I have used the online artificial intelligence tool VQGAN to produce visual art under the theme of truth, beauty and goodness. The AI is given text prompts and image models to begin generating images. I used the text prompts ‘truth, beauty, goodness’ with additional terms to create a variety of images. For the viewer’s reference, these text prompts are included in the slideshow. To enhance the experience, I combined the AI visual art with an AI-generated soundscape from Soundraw. My role has been to produce and curate the art.

This project raises interesting questions:

  • Is this art?
  • What connections do you see between the initial text prompts and the final artwork?
  • Do you consider this art interesting, attractive, or ugly?
  • As this AI ‘learns’ from human data sets it can be considered as a mirror of our humanity. What does this art tell us abot ourselves?

Do you have a favourite artwork? Do you really hate a piece? Why not vote to gain a collective perspective of people’s response.

You can see the summary of the votes here

About the Artist

VQGAN + Clip (from Open AI) is neural network architecture capable of generating images from text prompts. The models used are from Flicker images and imagenet.

About the Curator

Jason Hoet is interested in multimedia art and technology and has curious questions. This project is one instance of these things converging. Jason is also married to one, a dad of four, and Pastor by day (and night).