The image above was created by an AI, and won a digital art competition in the US, sparking controversy and fury among artists.
The artist went on to sell his winning art piece for over 700$, in the first instance where he worn the prize it’s still unclear whether he broke the rules of the competition or not, but during his submission he was asked by the judges what his process looked like in creating the digital piece and said, he used his computer and mid journey.
Then his submission was accepted in the contest, it’s not clear if the judges understood what he meant by mid journey, as AI generated art work was a new thing and it’s understandable for the judges to not have heard of Midjourney during that time.
After word went out of what had happened the Colorado State Fair, who were the organizers of the competition sent out this tweet.
These are two real life instances so far, where art that would normally be created by humans was done by an AI and commanded the same results, it won a competition and sold as art for a good price.
To be fair, the artist did not just open up Midjouney and on their first try generated the winning piece, he claims to have spent over 100hrs tweaking his prompts until he generated the art, but with AI getting better and better how long before you can generate a masterpiece in just 1 or 3 prompts instead of hrs of trials and tweaking.
This has led me to the question, what Jobs will AI replace in the 3d industry.
To answer this question we need to look at the production pipeline of a visual effects heavy film like Avengers Infinity war or SpiderMan.
How Ai will affect, story writers/script writer jobs
Any movie starts with a story that turns into a script, this is one of the safest jobs in the industry, safe at least from AI generators for now.
This however, does not mean there are no Ais generators trying to write movie scripts right now, today there at least 10 Ais script writers used for writing Ad copy, short stories and more. What i think is going to take some time is, triple A studios trusting an AI to write a compelling story that can be turned into a feature length movie or animation.
How Ai will affect Concept Art and storyboards artists jobs
Concept art seems to be the easiest entry point for Ai image generators, almost all Ai image generators released this year can be used for some form of concept art, and one of their best selling point is how quickly they can turn around different concepts and art styles.
Generative Ai has been compared to synthesizers that people thought would replace musicians and guitar players because with synthesizers people thought you would no longer need a piano player, just add the notes and the computer will play for you.
Synthesizers did not end up replacing guitar player or piano players, they likely made them more valuable, people want to see other people perform and appreciate talent, that’s why you go to concerts, endure the long lines, pay expensive tickets and spend hours standing only to listen to a song you can easily listen to, in the comfort of your home on your laptop or phone.
But Generative Ai is different from synthesizers, there are not many people who will sit to watch an artist create concept art on their computer for hours, most ordinary people just care about the final piece rather than the process of creating the art piece.
AI progress is too fast and too good, 90% of people can not tell the difference between ai generated work and artist created work, a few weeks ago each image you generated was slightly different with a slightly different visual style there was no consistency, now these generators are generating more consistency and can redraw the same character in different poses, something they struggled with a few weeks ago.
Where Ai may struggle in concept Art is i believe is telling a comprehensive story with consistency and continuity, yes the ai can generate consistent visual style, but what of continuity, if one image had an explosion and you have rocks everywhere would the Ai be able to place the same rocks in the next image it generates or would it need an artist to do that.
How Ai will affect Character design jobs:
This can fall under concept art, the same Ai generators we have today can be used for character design.
How Ai will affect Modeling and sculpting jobs:
As of writing this, there are at least 3 Ai generators capable of generating 3d models from text prompts, the process of generating 3d models has always been of sweat and hard work, this guaranteed jobs for 3d artists as its a field that always require human labor, we did see a movement towards procedural generators, though it was reserved to already time consuming tasks that don’t require a lot of artistic direction in the early stages of designs like, tree generation using speed tree, city generators using geometry nodes, environment generators like true train and more.
Now we are seeing Ai generators capable of producing models that were too complicated for procedural workflows.
Ai generators like Nvidia’s get 3d and stable diffusion can produce cars, bikes, animals all from text, the resolution and visual fidelity of these models is still low and unusable in most cases but this is just the first iterations.
Afew papers down the road and maybe there will be no need for places like cgtrader or turbosquid which are dedicated to selling 3d models. I hope that day would never come, because most artists make there money from selling asset packs and 3d models which these Ai would do away with.
How Ai will affect Rigging jobs:
Rigging being one of the most complex part of any 3d pipeline, i think this is going to remain a job for humans for now, characters especially animals are too unique and different to be easily rigged by an AI, rigging i believe is the last job that an Ai is going to replace, yes the Ai may be able to generate a standard rig that can be used for human characters but to create a rig for every possible animation character is just too difficult. There are so may unique poses for different character designs that an ai may not be cable of matching.
How Ai will affect Animation jobs:
As a result of rigs being too hard for Ai to generate especially for non organic object, i know i said animals are harder but now that i think of it, non organic objects that need to be animated require more specialized rigs than animals, a rig for a giraffe needs slight tweaking to work on a dog, but a rig for optimus prime is simply unusable for the pixar lamp character.
And for that specific reason, because each character has a special rig, that means the Ai would have to be tweaked to handle that specific rig for animation, which may take more time than just hiring a rigger and animator to do it for you.
Generative Ais are still new and for now, most studios are going to maintain their workflow and pipeline for the same reason why studios refuse to adapt new software like blender. They take longer to adopt new workflows. its a reliable and its what everyone uses and is used to, so for now we are safe.