Flirting with an AI art generator (Part 1)
3 months ago I was in my room in front of the computer browsing my emails.
Suddenly a rush of excitement smacks my spine. I see in the Futurist Magazine newsletter that another AI art generator is available to test. Its name is Midjourney and you can think of it as the lesser known Dall-e or the Solange for Beyonce, if that makes sense.
So I embark on my journey to creating art using the written english language. I sign up and head to their Discord. Apparently Midjourney works from there.
I need to pause the story for a moment to appreciate the fact that most of our lives, creating art, consisted of countless hours of practice, understanding light, color theory, perspective, proportion, rule of thirds, depth of field, and a long etcetera before being able to come up with a decent piece of art.
Now, adding a phrase with some key instructions using a bit imagination and voila, you have a decent (digital) art piece.
The barrier to entry has become lower. And the immediate consequences are maddening established artist and an array of unpredictable outcomes (more of this later).
Highly allured by the premise of being able to generate anything I want, I stretch my fingers and… I start reading the instructions. I don’t know where to start, so I guess that is the best way.
Ok. Now I’m ready to pour my imagination onto the keyboard. I write the command “/imagine prompt cryptocoin as a toon character “
Four dark squares and a zero next to a percentage sign appear on the chat I’m having with the Midjourney robot. The rendering starts, I see the squares change color before my eyes, and the percentage goes up. My mind is trying to predict the result, I imagine a coin similar to Pikachu or other cute manga character, or perhaps a Hanna Barbera looking coin but what if it looks clippy?. (20% rendered) Why does the screen look so dark? (50% rendered) it is just pixelated it will get better. (70% rendered) I’m not sure what I’m looking at. (100% rendered) What the…that’s not what I expected.
I have mixed feelings about creating art through AI ar generators. Part of me is really excited about this point in history where art and design is changing to become something new, similar to when photography entered the scene of art and art had to adapt and become not a portrait of daily life or the bragging shot of a fancy aristocracy. I’m excited about the possibilities this technology offers. With this new paradigm there are few barriers: your imagination and the data the AI uses to create the art.
That means less time generating decent looking ideas, or having to hire someone to create the art for you, or having to explain the concept for them, among others.
The other part of me was frustrated, Midjourney didn’t understand my query. On top of that I felt sad for the traditional artists and the digital artists. Their work is in a sense commoditized. And finally I was slightly confused. I wondered what does the AI art generation mean for artists? if I use the style of an artist (dead or alive), is that legit? How will art change based on this disruption? What would be the role of artists? What is the downside of not jumping in this wave?
By the way, this was the result of my query on Mid Journey.
Part 2 of the article coming soon.