What are some of the differences you see between these two images? For starters, the cars are not the same type, the orientation of the cars is different, and the one on the bottom looks like a photo, while the one on the top is a drawing. One could also point out similarities: they are both grayscale images, they both feature sporty-looking cars, they are both set in the woods, and both of the cars have a lot of detail reflecting the wooded environment around them. The most important distinction between these two images, however, is perhaps less obvious: authorship.
The image on top is a pencil drawing done by a human, and the one on the bottom was generated by a photoshop-based artificial intelligence (AI) program. The idea of doing this little human-vs-AI drawing comparison occurred to me the other day, when I read a breaking story on the use of AI by a well-known magazine. As the story goes, it was discovered that Sports Illustrated was publishing content generated by AI, and had been attempting to pass it off as created by humans, right down to the AI-generated (fake) author bios and headshot photos (link). When this was discovered, Sports Illustrated quickly deleted the content in question and blamed a third party vendor for the fake authors and related material.
It goes without saying that this sort of practice has serious ethical implications within the media, and should (rightly) cause even the most casual reader to critically ask where the content they are consuming is coming from. In addition to that, it is yet another reminder that different forms of AI, for better or worse, are making their way further into our lives and businesses seemingly by the day. At this point, it seems safe to say that the question isn’t whether or not this is a good thing – AI is here, and it’s not going away. One can only hope that incidents like the one with Sports Illustrated will generate discourse and the development of ethical standards on the use of AI in the media and in business, especially with regard to transparency in disclosure to consumers on where and how it is being used.
Returning to the above images, my comparison exercise quickly revealed one reason why the use of AI is so attractive to businesses – it’s fast. Really fast. I’m not sure exactly how long it took me to complete the pencil drawing shown above (I worked on it periodically over a couple of weeks), but I would conservatively estimate at least twenty hours. The AI-generated image, by comparison, was completed in about fifteen seconds. Not only did it complete the above image that fast, it completed several variations of the same image, offering me the option to choose which one I liked the best.
When I first considered the results of my experiment, I kind of wanted to slide my drawing pencils off of my desk and into the trash. Why should I bother to spend hours creating art when I can have my junky old laptop generate a high-quality image like this in a just few seconds? The next thought that occurred to me was another question – what will this emergence of AI into our daily lives do to us as humans? What will it do to our brains, our ability to think, our ability to independently problem-solve, our ability to create, to imagine?
As with many challenging questions, the answer is complex and in many ways unknown. At the same time, it seems to me like it will, in many cases, be a question of choices. When I was a child, if I wanted an image of my horse colored pink and riding a unicycle while wearing a football helmet with a unicorn horn attached to it, just about the only option would have been to draw it myself with crayons or the like. Now, however, my AI program could do that for me in a few seconds. Or a horse with frog legs. Or a frog with horse legs. Just ask the machine, and boom – done. Sure, I could draw it myself, but I wouldn’t have to. I can’t help but think this will affect the way we think, imagine, and create in significant ways. Besides that, this reality seemingly leads back to a question often revisited on this blog – why do we do the things we do? It seems that the answer to that question will play a role in determining how we use AI in art and other endeavors.
So, did I throw my pencils away after my little experiment? No, I didn’t. Will AI-generated art stop me from drawing? No, it won’t. I don’t draw for recognition, or for likes on social media, or to make a name for myself, and certainly not for money. I do it because I want to, because I enjoy it, because it’s good for my mind, and because it’s a cool personal challenge. In other words, I draw for the sake of drawing, and for me, that is a worthy end in itself. Perhaps in that is ultimately where we find our answer; creating art for its own sake is a uniquely human pursuit, and while AI may influence the how of doing art, it won’t change the why.
AB20

