Alright, so for the Middlebury art assignment I went with a recognizable person so it would have lots of reference material. I also determined that the person would have to also be someone who has already become sort of a caricature of themselves in pop culture. I chose Richard Nixon. Despise the man, but I feel like his face is quite unique and AI would have some fun with it.
The two sites I used were Craiyon which was suggested by the site, and 123RF which I had found on my own. Both are free, but 123RF limits the amount you can do in a day, and you have to make an account, so the fun is short lived and a bit tedious.
The prompt I started with was Richard Nixon. Below are the two pictures from Craiyon and 123RF respectively.
So, without any prompt other than a name, 123Rf decided to put him in a library with a book with what looks like his own fac on it? That sems Nixony. The Craiyon one went more cartoony so maybe it was pulling from caricatures instead of photos? either way, results are pretty simple.
I moved on to add some extra words. the new prompt is Richard Nixon signing documents. I didn’t want to jump up to crazy yet because I wanted to see if simple additions got wild. Below are the results.
In this iteration, its sems that 123 was able to keep the somewhat photorealistic style and managed to make sure all of his parts were normal. It put him at a desk and actually doing the prompt. Very nice. Now the one from Craiyon is unsettling. They switched it to black and white from cartoon so clearly in the caricatures he has never signed anything, only on TV. But, for the love of God, what is wrong with his hands. The pen has fused with his finger and his other hand is like… a flipper… I guess… I did not think it would get this bad this fast but, that’s what I was secretly hoping for.
For the last prompt we were supposed to add something crazy or in a style etc. I thought location would be more hilarious. Initially I said on a plane or in a pool, but I wanted it to have to think a little bit. I decided the new prompt would be Richard Nixon Signing Documents in a flaming golf cart. The results:
So the image from Craiyon is no longer showing me a recognizable Nixon, just a generic man in a suit inside a burning vehicle. It could barely be recognized as a golf cart so I kind of think this one missed the mark. It also went back to cartoon style for some reason… interesting. But the 123 picture is comedy gold. Why does h now look like Josh Brolin and why is his pen now a gun? Who are the faceless people? AND WHY IS HE STILL AT A DESK?
Which image was the worst? – I would have to say the last one from Craiyon because it started losing details and Nixon as not recognizable anymore.
Which one was most impressive? – In the comedy department, probably the last 123 picture. The desk on a cart with a pen gun is pretty amazing. But with prompt accuracy, I would say the second 123 pic because it didn’t really have any strange things about it.
Where do you perceive AI image generation as a threat to artists? – I don’t need to perceive it, it’s already happening. So, I am a huge nerd and play Magic the Gathering. This article came out earlier this year and it is but a small step to these artists losing their jobs. I mean, one of my favorite parts of the cards is admiring the art. Having the artists name on the bottom of the card lets me know that a person with this amazing skill is making the card game I love. I can then look them up and see what else they have done. I could then go buy their merchandise. Their art is their livelihood so i think it should be protected. When AI is used to make art for these cards, it really will take that small appreciation out of the equation. Would it lead me to stop playing, probably not if I’m honest, but it will feel less alive.