When I think of AI art, I don’t think of our machine overlords coming to take me or my creativity away. I think of it as another human achievement we should celebrate. A single AI constructed image is the product of thousands of years of human creativity. If I were to use hyperbole, I might even say it is the ultimate human accomplishment in artistic excellence. But I am not going to use hyperbole because well, the fingers are just not right. Especially if I am using the free AI image generators. Sometimes there are six fingers on one hand, four on another. Sometimes there are three lips or a couple of extra nostrils. Here, an arm is coming out of a thigh. There, an elbow is bends unnaturally. And sometimes there are creepy blobs of flesh where there shouldn’t be. I should say instead that this culmination of human creativity sometimes makes me feel slightly uneasy.
Long ago, I used a raytracing program that allowed me to generate images, not with natural language prompts but with code that I had to learn. The code hid the underlying graphics algorithms from me. Code that let me create circles and squares and camera angles and light sources and colors. It was fun. It felt innovative and groundbreaking.
There were contests. I still use some of the winning images on my laptop as wallpaper (none of which were mine). Where I spent hours, sometimes days programming a scene and gushing dopamine over whatever I managed to produce, I can now do in minutes with natural language prompts. Mostly. The prompts have to be very specific and clear to really capture what’s in my imagination. AI rarely captures my inner vision. Though it can surprise and delight me. It can make me see that the culmination of humanity’s artistic accomplishments exceeds my imagination or puny talent. Will this technology kill imagination? Or will individual imagination, talent, and creativity be redefined?
If I create an image with clever prompts and in a style that suits me, it can feel as though I have accomplished something. But, of course, I haven’t. AI is just filling pixels with colour chosen according to a set of algorithms and statistical formulas I have trouble understanding. Some of the images are truly evocative. Others seem filled with human emotion. Most of the images do not come close to what I had in mind. When I explore other users’ creation, sometimes I linger with admiration and appreciation for what AI has generated. Filled with anticipation, sometimes it’s like wandering around an art gallery not knowing what I’m going to find in the next room.
AI image art shines when it is generating the unnatural, the supernatural, the dark, the mythical, and the otherworldly. Though I follow some dark art accounts on Instagram, AI images match or exceed the strangeness and horror of some of the man-made dark art I have been appreciating up to now.
I have been using the Nightcafe site now for a couple of years. At first, the images it produced were frustratingly awful. Inconsistent, unnatural, and creepy. The images being generated in 2024 are better quality than ever. In some cases, it is very difficult for the untrained eye to detect that they are AI. Looking at the images of AI landscapes, animals, or still life I always look closely. In 2024, something will be wrong or off. It’s the same with the people. My god. Those fingers.
I am being used by Nightcafe (just as I am being used by OpenAI), to tune the results. I have no doubt that the comments, prompts and images I provide are used to improve and modify the models. I am surprised therefore, by the volume of comments from users that seem to attribute human creativity to the generated pieces. There is much congratulation, accolades and praise given to the apparent human creator of some of the best ones. Except the creator is in most cases only crafting a prompt (To be fair, some percentage of images are external or evolved by the AI based on an uploaded possibly original image). In my opinion, it doesn’t matter how long or complicated the prompt is, the resulting image is not a human creation. Currently, it doesn’t seem to matter how explicit I am about how many fingers a human being has. NightCafe’s free generators rarely get it right. By the miracles of technology, AI will, I have no doubt, eventually give us five anatomically correct fingers every time.
There are users challenges that are run on the site to determine winning images for a given subject, concept or category. It is really only the AI that is being judged. The incentive to participate is baffling. There is no question that my data is being used as a feedback mechanism to Nightcafe, but that is for another discussion. Another time.
I am very aware that all of this so-called art I create is not mine. Do I own the images my prompts create? I say: absolutely not. As a new phase of human creativity, humanity owns the images. At least for now. We are already using AI to make our jobs and life easier, not to mention more creative. Long form art, as we may soon call human-only made art, will one day be prized and remain the champagne of human effort. This is all so new that it will take a very long time for legislation to catch up on what to do with AI art. Who will hold the copyright to them? How will creations be monetized since monetizing everything that comes out of technology is inevitable?
Let me consider the value of human creativity versus the product of AI creativity. To be sure though, AI in 2024 cannot be defined as creative. Here I’ve tried to list some of my thoughts on the pros and cons of the AI revolution:
Cons:
Because we will not be able distinguish between human or machine generated artifacts, it will water down the standard for artistic excellence.
Machine generated art is more economical for propaganda purposes. The sheer volume of media that can be created by AI at a low cost could result in conflict at all levels of socio-economic strata.
There will be more reliance on the machine art which will steal artists’ jobs away.
There could be an explosion of human mediocrity such that what we used to consider art might become too expensive for the average person to enjoy or participate in.
Retraining the workforce to align with the needs of automation powered by AI, may not be possible for most people.
Pros:
AI has ushered in a new age a human creativity because it is, as of this writing, the culmination, the aggregation of all human art in the historical record.
We will use AI as a tool everywhere in our everyday life. We will accept that we won’t necessarily be able to distinguish between machine generated and human generated products. As art redefined, we will just assume that all art is either human or AI generated or a combination of the two.
These tools will make us better artists. We will collaborate with AI to build better art.
There could be an explosion of human ingenuity and innovation as we build on AI generated media (as I would argue humanity has always done with all technology).
Jobs will be redefined such that we will be able to adapt to new skills required to use AI for our collective or individual benefit.
No doubt AI will benefit humanity most when there is ethical collaboration between us and the machine. This technology is a door that we clearly cannot resist opening.
Addendum
I asked various GPT chatbots to give me the pros and cons of AI art. Each chatbot gave me some things to consider. I have edited and merged the chatbots answers. The resulting table is a collaboration between human and the AI. Here was my prompt: “Give me a table of pros and cons for using human creativity only versus using AI tools with and without human collaboration.”
