Recently I had a conversation with friends about whether AI generated art is “art”. Putting aside for the moment the disruption this new technology has brought, such as concerns over deception and labour displacement, if an artwork provokes an emotional response, if it genuinely moves you, is it “art” whether or not it is generated by a human hand?

As with anything, you get what you put into it, and my initial results with text-to-image software were abysmal. Although the image is not generated by a human hand, it is still a product of human imagination and the eloquence and precision with which one’s vision is relayed to the AI. So crafting the prompt is a skill in and of itself.

The U.S. Copyright office holds that AI-generated works are not eligible for copyright protection because they are produced by non-humans. Yet it is interesting that when giving attribution to these works we, humans, seem to consider it a joint effort. For example, “created by Mindjourney using prompts by Jason Allen”.

So I wonder. It seems that AI output still depends to a large extent on the personality, style, and vision of the artist. The text prompt of what is desired comes from the artist, and the ability to formulate the right prompt is arguably the mastery of yet another tool with which the artist can bring forth their vision. The final artwork - if it is good - still reflects the artist’s individual style.

And I also wonder, if formulating the right query is a prerequisite tool for expressing one’s vision through AI, are the boundaries between logic (methodical, deductive, conducted in accordance with a strict set of rules) and inspiration (the instinctive and emotional knowing that must be allowed to flow untouched by the confines of logic) becoming blurred? And if so, does erasing the boundary between the two release us from being confined to one box (logical) or the other (creative), therefore releasing the potential for greater creativity?

I don’t know. Can AI illustrate an idea? Yes. Can it, with talented prompting by talented humans, create an emotionally impactful image? Yes. It probably can even learn to create more of such images on its own but in the current state of technology at least, it does not have a unique vision of its own.

So is AI-generated art, “art”? And what is “art” anyway? Perhaps my answer to myself for now is much like Supreme Court Justice Stewart’s 1964 response when asked to describe pornography: “[I can’t describe it but] I know it when I see it”.