The first short AI prototype screencapture was widely covered by the media. Most of the articles are based on the blog-post I made:
It has been very interesting to observe the internet’s reaction to AI generated content like this. I would have never been exposed to so many opinions about AI if the post had not gone a little bit viral. I am eager to see where the whole AI thing goes.
The demo is a convincing look at how an actual artist can use generative programs as a tool in their process, and that’s what troubles me.
The PC Gamer comment section is full of back and forth wether it is OK to use AI art in an actual product. The discussion seems civil. Which is nice!
A theme begins to form: discussion on what AI will mean for us people. I think this is an important discussion to have. And it is very interesting to see where this all goes.
A renowned Game Designer and VFX Artist Jussi Kemppainen has recently showcased a new practical application of text-to-image generative AIs. Using Unity, Midjourney, and Stable Diffusion, the creator has managed to set up a playable 2.5D adventure game prototype with fully-generated visuals. According to the developer, the AIs helped him to save roughly two days of work on the character and three on the location.
In recent years, AI technology has advanced dramatically, and attempts to use AI in the production of illustrations and manga are underway, and the possibility of using AI in the production of more complex games is being explored. Jussi-Petteri Kemppainen , who has worked in the game industry for nearly 20 years, has reported the results of trying out “whether AI can be used to create 3D models of game characters.”
As we have told you in various reports, it seems that we are fully experiencing a historic moment where Artificial Intelligence will begin to be an essential part of our daily lives, being present in many of the technological advances that we use every day. But on the other hand and in addition to this, a situation has begun that we have only witnessed in works of science fiction such as the Matrix and its spin-off Animatrix, where humans take sides among those who are in favor of the use of this robotic intelligence. and those who are against.
While everyone is currently talking about the intelligent-looking and coding-capable ChatGPT dialog system when it comes to AI, the use of technology improved by machine learning can also be a blessing for the time spent developing graphics for games. This is demonstrated by a recent example by Finnish developer veteran Jussi-Petteri Kemppainen.
After journalists, marketers, graphic designers, and students, indie game developers are trying to use generative AI to speed up some parts of his game design process.
This mixture of generative AI and traditional game development tools has begun to see adoption in the mainstream game industry. From small indie developers using it to do tasks that were unavailable to them, to bigger developers using this method to add more elements to their already full worlds, generative AI is slowly changing the way studios approach game design. As games contain a large amount of art, dialogue, and music, they are sure to be one of the verticals that will be revolutionised with the rise of potent generative AI.
HackerNews was the main driving force of incoming traffic. Over 20 000 visitors came trough that referral. Funny enough, they did not seem to have any excerpt or video or image. Just a straight up link. That site must see tons of traffic. The reception there was very speculative. Quickly the discussion turned to what generative AI would mean to the industy.
Never occurred to me until now that a capable artist may just have a lot more jobs available to him once the AI-to-image revolution will be fully rolling. E.g. some kid will create a video game with AI art, sell tons of it and then need artists to make the updates. In other words, AI facilitates entry to the digital art world which leads to the need of artists. It’s like free samples.quadcore
In reply to another comment:
Yes but by artists you mean (artists – 2d artists). Because as a 2d artist that’s gonna hurt. And with these models sucking all the current total creativity, and styles, with a world where this is going to hit them, the overall new creativity and art created will decrease and AI copying previous styles and arts will start and increase.
And then maybe modeling will take a hit too?
I do think it’s cool thought. But ya I am not that happy with what I was saying. Could still be a good thing overall?? Who knows.nazka
The best artists will thrive, I am not so sure about the rest. It seems very likely that the simple graphic jobs will be taken over by AI.kukx
On Twitter the comments to post about the prototype have been mostly positive, sometimes overwhelmingly so.
To better gauge what people feel about the use of AI in a game, I set up a poll.
From the poll, it is pretty clear that people are somewhat divided on the matter. Most do not mind the use of AI, but some are dead set against it. I personally believe that in the end what matters is how the AI was used. And to what end?
The clip found it’s way over to Reddit as well. The post was pretty popular for a while and there was a lot of very good discussion on the topic in the comments.
Comments on reddit:
Oh shiiiiiiiiiiit, this is blowing my mind but also that is some workflow porn right there.SoysauceMafia
I freaking love how he integrated the “gibberish” text AI usually generates lol “Paicktpparrltk” my favourite Store !
made me laugh.Evonos
I find the whole “AI will destroy creativity” discussion fascinating when this actually opens up so many creative avenues to people who would otherwise just not have the resources to pursue their dream projects. Using AI assistance, I’m finally getting a webapp project I’ve had on my mind for ages off the ground, and making good-looking assets for games is finally no longer a prohibitively costly procedure. It’s not like there’s no skill involved either – just a different set of skills.monsterfurby
Looks like exacly the area where AI is usefull. Not to “steal art from real artists” but to help achieve dreams for those who has no money for good art but has passion to create stuff in different areas.Drakmour
Almost none of the reddit comments were negative, Must have been because of the post being on a Stable Diffusion Reddit.
Some new mentions of the project have appeared online. I will keep adding them here:
I myself am a 3D artist working in the games industry. I do not at all feel threatened by the generative AI. I have tried to stay on top of the current developments and have used most of the AI tools and it is not going to be replacing us humans anytime soon. Most of the work in current gen game dev is something that the AI simply can not do. Sure it can generate beautiful 2D art, but modern games are 3D. That means highly optimised meshes, shaders and level design that is very carefully directed. Things that are absolutely out of the scope for the current AI tools. AI feels like Google image search on steroids.
Mobile games are not going to be any different. UI elements, Sprite sheets, particle effects. They all need to be created by us human to match and make any sense. We can use AI to quickly prototype ideas. Get a starting point and then build on top of it for sure. If there are some aspects of the work the AI can do, those are usually welcome breathers for the most tedious and repetitive tasks or be rocket fuel for creativity.
That is not to say AI would be pointless. It absolutely will transform the way games are made, but it will act more like an assistant for the artists currently working in games, instead of a replacement.
But the point of this adventure game exercise is to see what would happen if someone really did try to make a game using as much AI as possible. To dance around its limitations. It means that this will be a game of compromises. Compromises that the player might not notice, but compromises that will drive the production in to directions we as creators might not have wanted to go.
The look of the characters, locations and items will be out of our control. We can choose what version of the generated images we want to use, but we have very little say in what the actual art is going to be.
There are also things that the AI is super bad at generating. There will be scenes and events that the AI simply refuses to do because of the lack of reference data. We will need to write those out of the srcript and think of something else. A human artist does not have any of these limitations
The AI truly is an extra member in the team. Not a tool, but an actual collaborator with strengths and weaknesses. Someone we need to factor in as a teammember. And that is at the same time exiting and very weird.
Absolutely freeing, but at the same time, very, very limiting.
DISCLAIMER: this post is only about using generative AI tools to aid game development, other industries have their own challenges to navigate.