Generative AI Art — DALL-E, Midjourney, and Stable Diffusion, Explained

Tracing the history of machine creativity

I shot scenes like this for years. Now I can generate it with AI. Credit: Nick St. Pierre (Fair Use)

A class of cutting-edge machine learning techniques have become incredibly adept at generating new images (and now, videos) from scratch or from existing ones.

This is called image synthesis and has become one of the most popular use cases for generative AI.

Update: I now post AI tutorials on YouTube.

Image synthesis has many applications, such as…

• Advertising — creating realistic images of products or environments
• Computer vision — generating synthetic images for model training
• Digital art — creating new graphics, paintings and photos
• Fashion — generating creative ideas for design and new products

Midjourney V5 — Photorealistic environmental portrait that never existed. Source: Christian Maté Grab (Fair Use)

Simply type some text prompts into DALL-E, Stable Diffusion or Midjourney and you can create photo-realistic images of people and things that don’t exist.

What a time to be alive.