AI art is not suitable in games

AI art may be messed up. I could stop writing here. Technology is already being misused by shady firms and people who would rather take the hard work of others than put in their own. There may be a future in which artificial intelligence may be utilized as an ongoing resource for painting, literature, and other creative endeavors, but the fundamental foundations are so shaky that achieving that aim seems unattainable. Already, good people are being screwed.

Squanch Games employed a machine-learning algorithm to produce graphic assets and even a vocal performance for High on Life, according to Justin Roiland. The shooter was launched earlier this week to lukewarm reviews, and knowing that key aspects of its stale universe took shortcuts like this doesn’t exactly make me want to play it. Roiland told Sky News that the Midjourney program was utilized to put “finishing touches” to the game rather than using widely accessible skills in the industry.

The usage of AI art in the game was quickly discovered by players thanks to some shady posters in the main character’s bedroom. They have a lot of general ideas for making assets based on well-known movies, like people and places that aren’t detailed enough to be called “man-made.” If you don’t know how AI art is made, it is made by feeding previous work into a machine-learning algorithm. The algorithm then uses the previous work and a set of prompts from the user to make a piece of art that sometimes looks like something a person would make.

Image 28

It’s becoming better and better, to the point that furious disputes are taking place online as the distinctions are blurred and stolen work feeds capitalist views of art as a commodity, which can only lead to dystopia. It’s all shady tech dudes behaving as if it’s best for the future, true Elon Musk types who won’t acknowledge an artist even if it means killing them.

Many of the assets and current materials required to develop AI artwork are often handed to these systems without the agreement of the artists, allowing these programs to become parasites on their skill, absorbing it into an expanding hivemind. I’m no technical whiz, but I don’t need to be to see that it’s a horrible concept that’s already pushing excellent artists out of work in a world where earning a living as an artist is difficult enough. As High on Life showed, Squanch Games cares about costs and would rather reach their goal with meaningless gibberish than with a human touch. Roiland’s rationale for adopting AI generation is predictable: hazy praise for an uncertain future and how it will make creative production much more accessible. Like his writing and performances, I believe the term he’s seeking is lazy.

Image 29

He argues it creates a “weird and alternative vision” of our reality, but that’s because it actually rips from it, attempting in vain to conjure up an intentional imitation. It is weird and alternate by design because it cannot be anything else. That alone isn’t enough to make it worthwhile, but the game has that delightful Rick & Morty comedy, poor writing, and some recognizable idiosyncrasies, so of course the internet is eating it up and pretending like this isn’t a big deal. This development will win us over if we get complacent.

Another big problem is that people outside of the echochamber don’t know how AI will eat away at the creative industries, putting talented people out of work and making AI the most common answer because it is easier and cheaper than human touch. A machine cannot feel, think, or create art that has a significant impact on humans since all it can do is replicate what already exists. Yet, as long as it serves a purpose, companies and creatives with enough influence, like Justin Roiland, don’t mind.

Image 30
Shopping cart close