Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

How do you justify using AI for your game art? I have thought about this, and couldn't justify it. I also know that I hold on to morals almost to a stubborn degree.

So instead I recently started learning Blender, texturing, drawing and while it takes time to get to where I want, I'm passionate enough to learn the craft instead of "cheating", as I unfortunately see it.

EDIT: To add, I don't put games which revolve around AI as a mechanic to this same category. They're a new thing and they're cool for exploring the frontiers.



> "cheating", as I unfortunately see it.

This is the product of the time. I'm old enough to remember digital painting, 3D CGI, and digital matte painting/photobashing being fought and ridiculed as cheating or impure. Not old enough to remember this about cameras, audio recording, and synths, but you can be sure they also were met with this.

> I recently started learning Blender, texturing, drawing

You make the end result, not the tool. Your message and rigorous execution is what matters. Of course you'll need the skill in visual arts and enough discipline for the attention to detail to be able to make something worthy, the tool alone won't supply all that for you, but diving into the established area helps a lot. As well as having the message to tell.

And of course you won't be able to create anything interesting using AI alone, just yet - it doesn't pass the quality bar, you'll be using it as a shortcut in certain steps.


> This is the product of the time. I'm old enough to remember digital painting, 3D CGI, and digital matte painting/photobashing being fought and ridiculed as cheating or impure. Not old enough to remember this about cameras, audio recording, and synths, but you can be sure they also were met with this.

Very good point, thanks. I need to think on this more.


I don't think it's as black and white as that. I'm not looking to take shortcuts, I'm looking to use the tools at my disposal to make a game that couldn't be otherwise possible.

On that note, the plan is to make certain icons & portraits on my own, in a consistent artstyle that's easy enough for me. Some assets will be purchased from traditional asset stores, while critical pieces of artwork such as the Steam capsule art will be commissioned. Chiefly I wanna use image generation for background images & pictures describing events - arguably not that important or up-front, so not a great use of artist time.


Yeah I agree that it's not as black as white, that's why I wanted to hear your justification for using it, to maybe update my thinking on it. And your use case definitely is one that is easier to justify in my brain than others perhaps.


How is

> How do you justify using AI for your game art? I have thought about this, and couldn't justify it. I also know that I hold on to morals almost to a stubborn degree.

different from

> How do you justify using Blender for your game art? I have thought about this, and couldn't justify it. I also know that I hold on to morals almost to a stubborn degree.

Aside from a vague implication that it is immoral to use an AI tool instead of <some other tool> to generate your own game art?


I've thought about this, and it's true Blender allows you to take a lot of shortcuts to be creative, like displacement based on generated noise. It has made work faster, and probably cut some low paying modeling jobs, because those automated shortcuts it has do things better and faster.

The distinction is that it's a widely accepted tool, and you still need to devote time to learn to use it good enough for you to be able to make satisfying art. It's like a brush for a traditional painter, only in 3d. In another comment below I explained a bit more why I value the amount of work seen in art.


I think an important distinction in good or poor quality AI art is similar to good or poor quality models made in Blender. A modeler can spend hours or days on a model; testing different cuts and resolutions and extrusions etc to get it just right. Similarly, the difference for AI generated assets will come down to how much time the prompt engineer will spend tuning their prompts, looking into different models, possibly training models for specific styles and outcomes, regenerating over and over again to get things just right.

On the other side, you have the modeler who barely sticks six rectangles together as a “human” and the AI art prompter who ships the first image they get back.

To me, neither is a moral issue. They both require using tools effectively.


For my game I am generating thousands of distinct level elements. I cut them out and rearrange them into level terrain and backgrounds. There is definitely a craft to this kind of work: figuring out what you want, creating the prompts, photoshopping the assets, composing the elements, etc. You could say that I am cheating in the art department, but my role includes the programming, animations, game design, level design, sfx, etc, etc. I justify it because I want to create something great, and AI generated art at least makes that a possibility.


Using tools is cheating? Why would I need to justify it at all?


I see games as an art form, and I see that the creative decisions are in game design, like levels, puzzles, and mechanics but also in the art style. Using AI to skip over the creative process of creating a distinguished art style and assets doesn't appeal to me.

It comes down to the question of how much you value the amount of human work in an art piece. When something is easy to do automagically, it loses some part of its value - to me at least, and I know I'm not alone in this.

Think how easy it is to create renaissance style paintings now with AI. Do you value those as much as the original ones? It's also part of the reason why traditional film photography has started to trend again, because there are people who value the amount of human work that goes into framing and compositing when they can't be easily fixed in post process.

You might think different and it's ok, just wanted to understand your reasoning.


This is an interesting take, I read your earlier message and assumed the morality part was about the "AI generated" artwork being (IMO) a derivation of the unauthorised unpaid use of the copyrighted work of other human artists.

Traditional film photography involved a lot of darkroom tricks! You can definitely reframe a shot with an enlarger, and it's also where "dodging" and "burning" effects/tools got their names. And of course, photography itself has been thought of as cheating - "real artists paint".

I think using tools in general is OK, and we already seem to draw a line between art (the creativity) and craft (technically perfect oil painting) so that seems consistent with mainstream opinion. Using "AI" for stuff currently feels a lot more like curating than making though.


TIL about those darkroom tricks!

Thanks for bringing the distinction between art and craft, it's a good way to rephrase my points. My point indeed is in valuing the craftmanship in addition to the art. And with current AI generators, seeing "AI used" label in a product immediately lowers my expectations towards the craftmanship of that product.

Also I'm building a traditional game with some new mechanics, so knowing that there are shortcuts now to building games like these makes me uneasy, as the craft of what was once appreciated can now be undervalued. Though I probably just need to swallow the truth.


AI isn't some sort of devil that takes your soul and magically makes your game successful. If you shoehorn AI art into your game without any knowledge of art, you're gonna end up with generic AI art. And if people buy it, then honestly, that's on them.

I think instead of crusading for/against the technology, you should look at what it can/can not do for you and then make a rational decision. For example, I regularly use AI to generate boilerplate. Not because I can't write it myself but because I've done it thousands of times already. On the other hand, I hardly ever use AI when dealing with obscure libs, because it's easier to just read the documentation than trying to debug hallucinations.


I think that any new tool just changes what's possible. At the start, we're all just rehashing the things we used to make, but faster (and possibly lower quality). This is the part that's scary for current practitioners, because the tools now do something basically for free that used to require a lot of human skill and it devalues our current skill set.

However, at some point there are people who really master the new tool and open up an entirely new range of possibilities because of what it can do. The value of craftsmanship just changes as the tools develop, it doesn't end. I don't know what new things will become possible with AI, but I'm confident that there are people with vision out there who will raise the bar on what can be created now that it's a thing.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: