Using AI for artwork

John is right: Speed is the main reason for using AI for artwork.

Before AI, most artwork was created using existing stock images or graphics that were edited to achieve the desired composition. Using AI isn’t much different. In essence, you’re generating your own “stock images”. Creating an entire piece from scratch usually takes too long. However, I still put manual labor into all my artwork. For example, I always add the text by hand and play around with details like color and texture.

I still enjoy creating pieces 100% by hand, but that’s typically reserved for themes like Halloween, Christmas, Mother’s Day, etc., where I can prepare something in advance.

The Monkey piece for 1710 was made with an imperfect AI-generated image, but that actually worked really well for the subject. The Whoopee Cushion from 1708 was also AI-generated because I couldn’t find a high-quality image of a blank cushion. I added the flappy "fart outlet" part by hand since the AI generated some kind of small label in its place. The jack-o-lantern was generated separately with AI and manually added on top of the cushion. The lettering (including the warping) was completely done by hand.

In my opinion, using AI isn’t much different from the "old ways." It may be actually more challenging because, in addition to knowing how to use Photoshop and Illustrator and creating good compositions, you now also need to know how to craft precise prompts to get the image you want.