Why Understanding Art History is Crucial in the Days of Generative AI

In this article, I want to explain why I believe learning art history is even more important these days due to the advent of commercially available generative AI tools (e.g., Midjourney, Leonardo, Stable Diffusion, etc.).

Midjourney has made it so easy to let our imagination run wild. But most of us know that harnessing prompts, style references, and character references can make or break the output. Although a simple prompt of “the brown fox jumps over the lazy dog” may yield great results, you may also get something like the following image.

I prefer my fox and dog to team up for adventures —

We used to say that “it’s all in the prompt”. But these days adding the right style reference can turn an OK output into something phenomenal. That’s why learning art history, artists, photographers, and their respective works is so crucial to the development of prompt engineers.

Let’s take the artist Hiroshi Sugimoto, a Japanese photographer and architect. I’m a huge fan of his black-and-white photographs (see below).

https://www.sugimotohiroshi.com/new-page-44

https://www.sugimotohiroshi.com/new-page-50

Knowing the artists and their works can vastly improve your generative breakthroughs. Although you could ask ChatGPT for the “‘best’ black and white photographer,” knowing the underlying art will lead you to better discoveries.

Now I can create something like this -

At the end of the day, I believe it’s important to know the rules before you break them. Stated differently, understanding the history of art will give you insight into why certain things were done, and how they were accomplished, with the result being you can pick and choose elements and toss the rules out the window.