Skip to main content

Indie game storeFree gamesFun gamesHorror games
Game developmentAssetsComics
SalesBundles
Jobs
TagsGame Engines

Will there be a way to separate AI generated games?

A topic by Morph created 28 days ago Views: 880 Replies: 12
Viewing posts 1 to 4

It would be good to separate games made entirely (or 90%+) by humans from the games made largely by LLMs, even if it was an opt-in category. This would allow human devs to better compare their games and connect with other humans devs, whereas vibe coders could connect to compare their techniques.

Jam Host(+1)

Yeah, this year there'll be a place on the submission form to indicate whether AI was used. I was originally thinking of making it a simple yes/no question, but I hadn't considered distinguishing between no AI usage, a little AI assistance, and games built largely or entirely with AI. That's a good suggestion! I'll look into making the options a bit more granular so people can better understand how each game was made. 

(1 edit) (+2)

Any chance the AI choice can be added to the filters? Otherwise with thousands of entries it will be impossible to find the kind-of games that you might want to rate.


Jam Host

That's a great idea! Here's what we've added to the submission form this year:

  • How much AI was used: None, some, most, or all.
  • What AI was used for: Code, art, audio, and other aspects of development.

Hopefully this gives everyone a better idea of how each game was made. itch.io has some support for filtering on this; here's what it should look like when it comes time to rate...

I was wondering about this too because it becomes difficult to define. AI is very helpful at the start for structuring the project, allow you to focus on the meat of the game logics, what % would that assist count? essentially advanced boilerplate. Also researching prerequisites going through documentation for you, answering individual element implementation question with code output, debugging, and generating edge case tests for you that you will prompt/plan out but still AI generated none the less. How much of these are permitted and technically non-infringing in a developer's natural creative input for the game.

(+1)

want to piggybag on this - your distinction is spot on and I would differentiate:

- the craftsman (all code handwritten and proud of it)

- the time crunched hobby dev (loves to code, architects the game, but lets the AI fill some cumbersome details) <- me, btw

- the game designer (has good ideas for great games but lacks the coding skills and thus heavily relies on AI)

Furthermore I would query this for assets as well. Which would give the reviewers an indication on how to evaluate the effort in a 2-dimensional AI space. So everyone could make his own rules/informeed decisions on whether it's a game they are considering reviewing or not.

(+1)

has good ideas for great games but lacks the coding skills and thus heavily relies on AI

How about someone who has lots of coding skills but sees no point doing things by hand anymore?

hehe, I guess I am oscillating between that and what I flagged :)

I like the IDEA of varying level of AI usage. I'm all for using AI to help with the heavy lifting portions but I am completely against it in the design and creation areas. 

(+1)

Without a way to granularly set your AI usage, and a way to filter it when reviewing, this jam is going to be a hard pass for anyone in the "craftsman" category. It's very likely this jam gets overrun by low-effort generated games, and everyone slings around "slop".

Less time spent doesn’t necessarily mean “slop.” With the right concept and AI tools, I can make a solid game in a day. The effort shifts toward vision, design, and iteration rather than doing all the coding or drawing by hand.

I meant that everyone slings the word "slop" around, whether the games warrant it or not.

Incidentally I don't see how you can have enough vision or design or iterations in one day. But good luck to you.

(5 edits)

Not exactly in one day for all of that. I may need a few days to come up with a concept, but once I have one, one day may be enough. In fact, I don’t know what I need a whole month for. The theme is generous, I may even submit two or three games.