I do a lot of experimentation with trendy AI generative art apps such as Dall-E and Stable Diffusion. These allow you do specify a picture of what you want, and will generate something they think it’s statistically likely you’re asking for. I see it as a tool: I take strong exception to the idea that the art is made by the AI, and not by the user using the AI as tool. I believe criticism of the provenance or artistic merit is similar to criticism first leveled by painters at photography: "That's not art, it's just pointing at something and pushing a button." Eventually generative AIs will be as unquestionably accepted as an artistic tool as cameras now are.
I’ve always been interested in using algorithms to augment my art and even as collaborative partners, both visual and musical, since working in 1987 with composer Joel Chadabe, an early pioneer in algorithmic composition and founder with Robert Moog of the first electronic music studio in the United States, for whom I ran SUNY Albany’s later Mac-based electronic music studio for two years, and worked with at his algorithmic composition software company, Intelligent Music. As such, have been a participant in the use of algorithms in artmaking for over 35 years.
Even before the boom in AI in recent years, I was writing and using algorithmic and generative apps reinterpret and expand my ideas in both music and visual art. However, the recent boom in Generative Adversarial Networks has made it easier than ever to produce very convincing and complex images. I’ve had a blast playing with them in the last year or two.
