Is AI the End of Human Creativity?

 If you're actively interested in art and tech, you've probably heard a whole bunch of alarmist wailing about the newest generations of art AIs. MidJourney, Dall-E, and others, are stirring up this debate of what art is, what tools an artist can use, what constitutes ownership, and what the ethics vs. legality say.

You can probably tell I'm itching to spit out my spicy take on this one, so I'll go ahead and get it off my tongue: this "debate" is mostly clickbait psuedo-panic. OF COURSE auto-generated art is not going to "replace" human artists. Did cameras replace painters? NO. This is a "yes, and" situation. 

The tools also aren't nearly as easy to use as the melodrama suggests. I'm currently subscribed to MidJourney, and loving it, but my results are about 80% failures. For example, I wanted to see a lumberjack made of wood, standing in a forest made of meat, because I'm weird. This is what it came up with:

Kinda interesting, for sure, but not really what I had in mind.

I've seen some people achieve incredible-looking results, but I've also seen the prompts they typed in to achieve them, and they were not plain text. They're paragraphs of text, code, parameters, cues for the machine. You have to speak AI. And that's a skill, in my opinion.

Personally, my favorite results from playing around were when I gave the AI a broad prompt and then just let it do its random thing, with no specific end goal in mind, just letting curiosity lead me. 

This is four of the variations on the prompt "opalescent floral fish."

Artificially generated art can't end human creativity. It can't exist without human creativity. Even in my limited experience with the AI, I've been able to see the bones of what it's drawing from. The machine learning only has anything to learn from because of art. And after all, isn't it human creativity that drives us to create tools like this?

As a writer with aphantasia and no significant artistic ability, it has been amazing to be able to put in some of the things I imagine and have actual images to attach to those abstract concepts. Isn't there a certain amount of art to that? Would everyone equally think to WANT a picture of opalescent floral fish? Maybe that's wishful thinking, because I love visual art and I've never come close to creating the things I imagine before. For me, this feels like a tool. Because I can't make pictures in my own brain, but I have a lot of concepts, I can put the concept into a picture-maker and then look at my results and narrow it down-- oh no, I thought I wanted that, but I don't, and actually this would work better in another color, and oh maybe this concept is more childish than I imagined it. And ultimately, I'll end up with something I could take to an artist to commission, or take to Photoshop to edit, or maybe even try sketching myself.

In fact, since subscribing to MidJourney, I've been doing art again. I've been sketching and learning to draw on the iPad and trying out various techniques. That feeling of possibility is the ultimate food of creativity, for me.

There are issues with AI art, of course. The question of consent from the original artists the program is learning from is an important one. The question of copyright and marketability, when it comes to things like using a MidJourney image as a book cover or promotional sticker. There are bugs to be worked out, as with anything. But we can't condemn AI for having a few areas of ambiguity, or for simply soaking up hundreds of years of art and spitting out someting similar; all artists do.