In the past year there has been considerable concern in the art world about AI art. As an artist whose income relies on producing human made works, I can totally understand the concerns. I feel like there are two main aspects to the issue:
There is the fear of our own style being stolen and replicated by the same said software, which has apparently happened to some big name artists still living and working and not just the likes of Monet, or Picasso. I know very little about legal matters, or computer software, but my understanding is that copyright laws allow works to be used in ‘research’ and that has enabled them to be loaded into AI data base to ‘train’ the software to make art in that style. There are AI works winning art competitions which is indeed alarming. Whose work is it actually? How much skill is involved in entering the prompt that AI uses to generate the piece and who should actually be receiving the credit? Since I am small potatoes in the art world, I have little concern of my ‘style’ being copied by AI, but that doesn’t mean we shouldn’t support figuring out what changes in our laws need to happen to keep up with the technology that has speed right past us.
That brings us to the second, and more encompassing fear of being replaced. I feel like digital artists are especially vulnerable at this point because the outcomes of the AI software that is copying them could conceivably be very hard to distinguish from their own work. Also it seems that graphic designers and illustrators would be at risk of what AI can currently produce, though at this point it may still need some human tweaking and input. But imagine if a company could hire just one human inputer and tweaker (job descriptions in this new world?) instead of ten artists making digital art on their tablets. Of course there would be a great deal of monetary savings involved, and unfortunately a loss of human jobs that would be reminiscent of the industrial revolution, or the employment of robotics in assembly lines.
But, rather than join the panic and scream doomsday, I would rather figure out what this means for me. AI is not going to go away no matter how much we stomp our feet in protest. It is here, and it is going to improve and become even more capable. Just as the advent of digital art did not spell death to actual paint brushes, I don’t feel like there will be an end to human art. We may need to pivot a bit, and adapt, but this is not the first time artists have done that… we are, after all, creative people! In my 61 years, I have seen lots of changes. I well remember when man had not yet stepped on the moon. I remember when simple calculators were the latest and greatest and when xerox machines became a thing. My family acquired a home computer when I was in high school… the first of any other family I knew! Our household didn’t have internet until halfway through raising our own children. And smart phones? Talk about changes! We can choose to fight them…. or accept them and figure out how they can be a useful tool. There will be growing pains. We need to figure out new legalities and logistics just as we have with every new rabbit hole humanity has raced down along the way.
To be clear, I still reach for a physical sketchbook when I have an idea, and usually only use photoshop or procreate to touch up or arrange my physical drawings. I am definitely not up to speed with technology. But I did decide to do a little checking out of AI apps to see what they are all about. I only tried a few, and the one I settled on experimenting with was Wombo Dream, which is a free app with an option for a paid upgrade. I know there is even more capable software out there, but this seemed like a realistic and accessible starting place. It asks for a written prompt, after which you select a style (from several options) and you can even give it an image to be influenced by, so you can add in some of your own art or a photograph if you want.
I kept it simple and entered the prompt “a bluebird riding a bicycle”. Some were pretty wonky and not what I would consider very usable.
Other style choices hit the mark a little closer, but would still need some work… like this cartoon version which is actually a blue jay with one leg:
Or this one that is just plain kind of weird:
Or this anime style that is blue, and a nice image,… but not a bird!
But then, there were also some pretty decent ones like this:
Which for me raises the question of using these as idea generators, or even a reference image. Is it that different than using an image from Pixabay or Unsplash to refer to in art making? Neither of those had any images of a bluebird riding a bicycle by the way… If I wanted ideas for illustrating a blue bird riding a bicycle is this a viable tool? What about entering in my own art and getting some new ideas for it? I didn’t have a bluebird of my own handy, but I did enter this robin from some of my own collage art.
Here are a couple of different AI versions from that input:
I tried to get it, through the typed prompts, to put my robin on a bicycle, but didn’t have much luck. This is as close as I got, so it seems like there are more limitations when trying to use your own image.
But even that gives me the idea of putting a big bird on a little bitty bike. Hmmm. Give me your thoughts… is there a way for AI to be USEFUL to artists and not just a threat? How do we AI proof our art businesses? How do we need to change to accommodate a new reality? Have you tried out AI? I would love to hear your thoughts!
Leave a Reply