Ask HN: Is it worth avoiding AI while making a game?

In my opinion it depends on how AI is being used. In many projects it’s hard to avoid it completely for example, while building a game solver I ended up using AI in a few places. More or less half of the development was done through vibe-coding and for screenshot detection I trained a small object detection model by AI. That part is difficult to solve reliably with simple rules and logic alone because screenshots vary a lot between devices.

Here it is: https://www.aiblockblastsolver.com

At the same time, most of the logic, debugging and problem-solving still came down to regular development work. So using AI with clear limits feels more like using a tool than replacing the work itself.

If this is your first time making a game, you should do everything yourself the first time.

Doing all of the work yourself at least once (or at least really giving it a go) will help you develop two of the most important skills in game development: understanding what parts of the process of game dev you like and are best at, and understanding how much time and effort it takes to get to something you think is good.

AI shouldn't be used as a replacement for not learning a skill. You will never have as much control over an AI doing the work for you as you want, and your role as the middle man between whatever AI agent you are using to do all the work and the end result is not a good or fulfilling place to be.

You don't have to avoid AI all together, but use it in ways that enhance the skills you care about developing instead of filling in for ones you lack. A good example for art is using it for placeholder art when you are starting something new or experimenting with an artistic choice. Using AI as a more convenient way to look up information while you are coding is very different than prompting what you want done and blasting it directly into your project.

Do you want to learn how to draw or buy a 3D printer with a pencil holder attachment and download images for it to draw for you?

So I've handcoded and released a couple dozen games in the past, but I'm now old with responsibilities and a day job of coding, and my brain just doesn't have the energy to mess about with it like it used to after a day of work, so I have started experimenting with letting A.I. agent code some things (and been surprised at how well it can work).

I'm aware of the gen AI stigma but it's either that or this damn game probably never gets released at this point (I basically let the game sit dormant all year last year, only recently picking it back up again), and I have a full graveyard full of unreleased projects the past decade and would like to finally get something out there again.

That being said, at least for the forseeable future, I won't use any A.I. generated art or music in the game (unless I inadvertently hire an artist that does use it, I guess), because there is still a huge (and somewhat justifiable) stigma against A.I. generated art and music assets.

So instead I'm just doing what I did in the past, and creating the art and music myself as best I can with Illustrator or Logic Pro, and keeping my games fairly simple and abstract so it doesn't seem too much like programmer art (I might at some point pay an artist or musician to improve things just before it gets released though).

I would like to try game development, only for personal satisfaction, and I will definitely be using coding agents and AI generated assets to help. I have very little free time and I am a decent enough programmer that I think I should be able to pick up what the agent is doing if I keep the chunks small enough and review each time. And this will be for my own hobby purposes and possibly for my children only.

But, if I were to make a game for commercial purposes I would very much either avoid using AI, or use it only for the code (and not reveal that), because as you rightly point out, the indie gaming community is massively against AI, and will reject your game if they know you used it.

I 100% recommend that you avoid AI-generated assets in your game. The stigma around AI-generated assets is very real, and not going away soon. Although people might enjoy using AI for their own purposes, by and large they don't want to be subjected to other people's use of it. Moreover, while I don't have the data to back this up, I would have to think that among the segment of the population that plays indie games, the stigma around generative AI is even greater.

In the second part of the question you asked if you should just learn all of the skills...buddy, does that question not answer itself? Of course you should learn all of the skills. Obviously that's much easier said than done, but TBH I think the quality bar to producing something viable is not super high, so as long as you're not a perfectionist, you can probably do it.

Since I could be labeled as an "AI hater" based on those comments, I want to be clear that I'm saying all this to keep you from falling into a trap and not to further my own agenda. The generative AI route is not a magic shortcut to success, although it is being aggressively marketed as such. The shortcut only seems to lead to success if you ignore the fact that people don't want to be subjected to other people's AI content.

Depends on what you're using AI for. Programming purposes? Most people aren't too bothered by that all things considered. Making assets like art and music? Yeah, people really dislike that, and there's a huge stigma against games featuring such assets on platforms like Steam.

Even if you manage to ignore the ethics issues and controversy, AI generated assets have the issue that they can be difficult to work with for a game developer, since these tools don't really make it easy to keep everything consistent across a game with thousands of resources. For example, if you made a 2D platformer with a pixel art aesthetic, AI generated graphics probably wouldn't work that well there. They'd look generic, show various signs of bad pixel art practices that would make them stand out as unnatural, likely not match the style for other assets in the game, be difficult to modify while keeping the same look and feel, etc.

Creating one piece of artwork with generative AI (as ethically dubious/controversial as it is) is 'manageable', since you don't need to care much about stylistic consistency, edits, etc. Creating all the assets needed for a game with such a tool sounds like a nightmare.

The hypothetical you state only matters once you have a game! The biggest risk by far is not AI assets -- it's finishing the game.

So if AI increases your odds of finishing, go for it. Then once you have a game, more people will care about whether it's good than whether you used AI assets. I suspect there will be lots of interest in how you incorporated AI, maybe even moreso than otherwise. You could alternatively use the AI assets as placeholders and intend later to replace them with hand-drawn, if desired.

When I made games, I had zero interest in making assets but wanted to understand every detail about graphics engines. I just grabbed random mediocre assets from online. I would have definitely used AI to make assets but done the coding myself.

It's a transformative technology, you should definitely try it at a minimum, and long enough to get a feel for it. It takes time to learn what/where they are good/bad at and how to interact with them to get the most of out them.

I saw a GameDev talk on Ai where they showed a virtual pile of trash. It cost more than $10k, what if we go photograph trash piles and use Ai to turn them into assets?

The Steam label, maybe it means something now, but longer I think it fades. For me personally, if there is a good game than looks nice, I'm not really going to care how much Ai they used. Be mindful of where you derice industry sentiment from, and that sentiments are changing.

If you're targeting desktop, Steam will be your primary distribution channel by a huge margin. Steam is AI-hostile. Don't use it.

If you're targeting something like mobile or web, do whatever you like with AI. Some will avoid it, but they're a vocal minority.

If you don't use AI, making a genuinely good game will take multiple years.

I'm not sure how fast AI will speed you up, but many of the things that would take time (creating assets, engine optimization, balancing) could be aided by AI.

It's best to avoid using tools and innovations that maximize your productivity.