Hi, thanks for taking the time to write this up.

I would like to address your points here, because I think it's important and I'd like to clarify the stance for others who may be curious. Long response incoming :)

Malic and I did discuss this rule and he was a little more ambivalent about it, it was me who was pushing for it, mainly because there has been an uptick in AI fatigue and derision as of late. I'll go through your points one by one.

I'll preface this by saying my points aren't directed at you or your game, just the premise of using AI going forward in general. Don't take any of this as an attack on you or the projects you've submitted.

1) You can use human-created assets for your game, whether they were made by you or not. We want to see what you create (whether that includes stock assets or not), not what a machine generates for you. I understand that there is also some effort you have to go through to get that generated content into your game, but we feel that effort could be better spent making your own assets or finding assets other people have made. You can kitbash, mishmash, programmer art whatever you want and I'll still be happier that you went through the effort of doing that and learning, rather than got a computer to spit something out.

2) Genuinely, I would rather see a text or systems-based game that has effort put into it than something with AI generated visuals. I would rather see programmer art. I don't believe that AI significantly lowers the barrier of entry. There are plenty of game concepts and systems you can create and design to make a simple game that doesn't require the use of AI. As you said, you were the only person to submit a game that used AI in the last jam. There were other people who had never made a game before who didn't use it, when we didn't have the rule in place. I think a developer's time is better spent making or finding assets than collating and fixing AI generated content, or simply rethinking how to approach a game so that it doesn't require its use.

The sensibilities of the organiser(s) are always going to come into play. You are for the most part free to make your game however you want, but every game jam has its limitations. Some have more, some have less. This one is quite open and isn't judged or ranked in any way. My thinking around this is that we want to see a game that YOU have made. We want you to challenge yourself. 

We also encourage you to work with others, you are free to collaborate with someone else who does have the skills you want. You don't have to do your projects alone -- if you are lacking in one area, I again encourage you to work with someone who can do that area. I reccommend anyone who wants to work collaboratively on a project to make a an open thread with what you're looking for!

This point is getting a bit long, so I'll move on.

3) You are correct in that, given the jam is not judged or ranked, your submission doesn't really affect other people's submissions. Even so, I don't particularly feel that AI generates interesting content. I would rather see programmer art that was made by a human than anything generated by AI even if it's visually stunning. I think people are fairly tired of the overload and overuse of AI content, especially within the last year. I've never read interesting AI-generated TF content, it all sounds the same. Formulaic. Sterile. Generic. When a person who is actually INTO transformation as a concept or trope, it becomes interesting because it is being written through the lens of a person who appreciates it, not through a black box of approximation.

4) I think the better way to frame this is that instead of using AI as a crutch, you should think about novel ways to make systems that don't rely on it. To use your previous submission as an example, you could make a modular sprite system and make a few variations of sprites for different scenarios, and create the rules for how they should combine. I would much rather see that, even if it ends up looking janky. Again, I want people to challenge themselves. It's up to you where you spend time on your project, and you should consider the amount of time and scope your project will need in various areas. There is ample time before the jam starts to come up with ideas to work within the constraints of the jam.

5) It is, at the end of the day, up to you if you want to make a project for the jam. I do want more people to make TF games, but I want to see what humans make, not what a machine makes. Again, you have plenty of time to come up with ways to make a game that works within the jam's rules.

6) I generally feel that the opinion around AI has shifted substantially since even a year ago, as has mine. When it was first cropping up, I enjoyed seeing the weird things it could make that I couldn't devote the time to doing myself. However, as time went on the issues with it increased, and the novelty decreased. My opinion, along with others, shifted. There is a lot more AI fatigue around as of late, people are tired of it. It's usually quite formulaic, and the sentiment is prevalent that it is just stolen content approximated into something 'new'. I would agree with many who are against the use of it due to the fact that it is trained on the work of artists that did not consent to have their work taken wholesale and put 1:1 into training data. Genuinely, just make programmer art instead. It can look bad, but at least it's your original content, or if you use stock or someone else's assets, it's another human's content.

7) I don't even know how to see the plays of other games in the jam. It wasn't a consideration, I'd forgotten about your submission since last year and the rule's addition was based on wanting to see things a human has made, and challenging participants to be innovative. If I had thought about it earlier, it would have been a rule in TF22 & TF23. It was not added because of your submission.

Again, I encourage you and others to challenge yourselves and collaborate with each other to fill in areas you are weak on, rather than relying on AI just to get a submission out there.

Cheers ^^

I am personally super happy with the decision to ban AI artwork from participation. I've used AI to fill in the gaps in some of my projects in the past, but I know for a fact my games were worse off overall for it. It robbed me of a chance to practice what I wasn't confident in. to folks who are determined that they have no artistic talent, just give it a shot! it may take you a while (my games are a prime example of my slow growth), but giving up at the starting line wont help you in the slightest. for a game jam focused on encouraging creativity and practice, this is the right call! good luck everyone! 

My question is basically where is "the line" when it comes to AI use? 

Is it against the rules to use a LLM for debugging or optimizing code? Seems like it should be given that LLMs are trained on countless lines of stolen code. But even then, how would you verify without critically analyzing the source code of every entry?  A lot of the same logic used to condemn image generators also applies to LLMs but  it's just not as convenient to ban those.  An artist making a game doesn't have to "git gud" at programming if they can use an LLM to do a huge chunk of the critical thinking and creative problem solving for them.  But as a programmer, I still have to learn all about perspective, vanishing points, proportions, color theory,  brushes, layers, blend modes, on and on just to have a bare minimum standard of quality and that's the rub.

I hate drawing. I hate it. Every second I spend drawing is like pulling teeth. Visual art is the least enjoyable part of the game dev process for me. I have spent more time trying to draw something I don't hate looking at than anyone could reasonably be expected to. And at this point, I don't even WANT to get good at drawing. I want to spend that time becoming exceptional at something I enjoy like coding and music rather than becoming mediocre at something I hate like drawing. This is why I'm building a typing game for my entry. It requires very little art and animation  so I can minimize the time I spend drawing and focus on adding accessibility features and polishing the experience.

So while I get the logic behind banning AI images, I don't really think it's consistent unless it's applied across the board for all uses of generative AI including debugging and optimization. Because from what I read, this jam is all about what YOU can do, not ChatGPT right?

I'm perfectly happy with this decision. We're competing on our skill, I'm almost single handedly making my game with help from a spriter friend, and a coder friend, using PAID FOR HUMAN MADE tilesets. There's genuine effort in every aspect of my submission. If I had some hackjob ai slop even come close to me and my friends' efforts I think I'd lose even more faith in humanity. I have STRONG feelings on this topic. I'd rather not have someone cheat on their homework and get graded on the same level, you get me? AI conceptualizing is fine, but if you incorporate that into a large portion of the game I will only ever see it as cheating and cheap. That's my piece.