为“在乎”辩护：IGF 评委谈游戏大奖为何需要明确的 AI 政策

In Defence Of Giving A Shit

10/11/2026

A few weeks ago I was one of the judges for the Independent Games Festival. Judging for the IGF is one of the highlights of my year, it's a real joy, but every year there is usually one discussion thread that gets a bit heated. You can normally tell which one it is because it has more comments than all the others - most might have 4-5 comments, but the spicy one will have 15. When I first saw the thread about Big Walk this year, it had 90 comments. I think it had 150 by the end.

Games this year had to have declarations if they used AI. Unlike Steam's current declaration policy, which only requires developers to declare if assets are "consumed by players", the IGF required a declaration for any and all AI use. The upshot of this is that a huge number of games declared their use of AI, including games made by prominent indie developers who have not declared their use in public. The most controversial case was Big Walk, whose developers stated they had used LLMs for "less than 0.1% of the code" to fix errors late in production.

The resulting discussion thread about this included some of the stupidest things I've ever seen anyone write about AI and game development, which truly is saying something. When it blew up today in public, I saw the same arguments being repeated. And I wanted to write this, not to rebut every single argument I've seen, as tempting as it is, but to talk a bit about the actual shape of this problem and what is actually happening. Because what's happening is not unusual, it's not extreme, it's not unfair and it's not a 'public flogging'. What's happening is exactly what is supposed to happen in creative communities. The argument surrounding Big Walk has two parts, so I'm going to break it down that way too: first, should the IGF have a policy on AI content; and second, should House House be disqualified from winning an award. And then I'm gonna ramble and yell a bit and then upload this post and go and lie down in a dark room again.

The IGF Needs an AI Policy

The IGF desperately needs a policy on AI content of some kind, whether that policy is for or against. The problem is that their existing (unwritten) policy was to allow individual judges/jury members to decide on what they wanted to do. I appreciate why you'd have a policy like this, it fits the subjective and personal nature of judging for arts festivals. But the problem is that the AI issue was happening so silently behind the scenes of the IGF that the average indie developer in 2026 is unaware that they are paying money to submit their game to an awards show where 1/3 of the judging pool are immediately going to abstain from considering their game. And I say that as one of those judges, like I'm not in favour of those games being in the IGF, but I think it's unfair to not clarify this massively important part of the judging for an awards show that you have to pay to enter.

But if we agree that it's only fair to add a policy, what should the policy be? For many judges, the idea of banning AI content of any kind - even art - made them mad. Many people were fine with banning it for things like art or music, but felt very strongly that AI code is simply the norm now and shouldn't or can't be banned. The reasons given for this were mostly very stupid, and part of me wanted to list them out here and break them down but I think both of us would lose brain cells in the process that could be lost in a more fun way like playing videogames or running repeatedly into a brick wall. There are so many arguments against this that I truly am losing count of them these days, we have just mostly agreed to ignore them one by one.

What I will say is that these discussions exhibited a lot of really, truly, deeply disingenuous arguments, and I see this more and more now. It was fascinating watching people who know absolutely nothing about AI talking confidently about what it can and can't do, or how it does or doesn't work. In many cases the writer was not even trying to be understood, they were trying to intimidate others into shutting up by wielding scientific/technological language as a way of appearing authoritative. I think this is one of the most hateful things you can do in tech or tech-adjacent spaces. I love science communication, and one of the things that makes me saddest is when people offer up their own stupidity before we've even begun talking about a difficult topic. And the reason they do it is because they spent a lot of their life being spoken to like this, by the worst kind of intellectual bully, and they've concluded that people around them must just understand these issues better.

Of all the arguments against a ban though, there's one I do want to address because I think it gets at the core of "should there be an AI policy", and that's the idea that bans don't work. People pointed out that if you ban AI, how will you enforce it? People can sneak things in any time. And they're absolutely right. How would we have ever known that House House used AI if they hadn't declared it? They could've just snuck it in and it would've taken a significant internal leak for it to ever come out. So maybe banning is pointless?

Here's the thing: when you write down a rule, it doesn't do anything. Rules are not magic. They don't have any power over people. But they do two important things despite this:

1. They tell everyone outside the community what to expect from that community's space/awards show.

2. They give everyone inside the community something to point to when a violation happens.

So the rule is both a signal and a justification. It tells everyone what you stand for, and it helps you clearly enforce what you stand for when it matters. This is how it works for PROCJAM too. You cannot enter PROCJAM with generative AI. If I see it, I will delete it. But can you sneak it past me? Of course! You do not need to be very clever to do this. But by having the rule we are saying: look, this is what this community stands for. This is what we will and won't accept. And that creates a sense of community and space. And a lot of people respond to this, because we naturally understand what communities are and what they are for. The same goes for award shows! Many award shows have requirements to meet, and they define what the awards are for and what they believe is important.

The other thing it does is outline consequences.

Although House House could have hidden their use of AI, would they have risked it if there was a rule in place? Probably not. Because the cost of the news coming out is amplified many, many times over if they violated an explicit rule or policy. So it creates a little bit more social pressure as well. Of course people will cheat, but usually they're the kind of people who already don't care. The people who are invested in the community are risking a lot more by doing so.

Bans on the use of AI should not be thought of as shaping reality or controlling what actually happens in spaces. They are about defining and empowering the community around them. This is actually the biggest issue the IGF faces, and the reason the debate was so intense this year: because the IGF no longer has a coherent single community, as evidenced by this issue. It is not really possible for the IGF to continue as it is, because defining a policy will instantly alienate a third of their judges no matter what. But I don't see another way forward, because the competition is currently a chaotic mix of opinions that I think, ultimately, will put people off submitting.

So whatever your position, I think the IGF desperately needs some kind of policy, but pretty much everyone involved is agreed that it doesn't have one now, so we have to address the situation we're in some other way. So:

Should House House Be Given Awards?

Whether or not the IGF should have an AI policy, the fact remains that it does not, so for this year we have to fall back on their current stance which is that individual judges and jurors get to decide for themselves. A lot of games declared AI use in the IGF this year. A few of them would surprise you, but probably none would surprise you more than Big Walk, which is the one that hit the headlines this week. The declaration submitted for Big Walk clarified that LLMs had been used to diagnose and fix some errors towards the end of development, although personally I still find the explanation a little hard to unpack. Regardless, this usage is probably close to the smallest possible amount of AI usage you will ever see in a game. So, what do we do.

The most immediate and obvious problem is that people liked playing Big Walk, and lots of indies know and like House House. If you read the response to the discourse on Bluesky, a lot of the responses boil down to "they are good people" or "they made a good game". I completely understand this response, it's very human and natural, and it's great to hear. If you're suggesting that exceptions should be made for this stuff if we agree the game is good, or if we are friends with the developers, that's also okay but I don't think anyone actually wants to commit to that as a position. How well you know House House should not matter, what should matter is whether you think they made a mistake. Some of the responses that disappointed me the most have made appeals to how good the people involved are, or about how responses to this should respect people's ability to grow or whatever. No-one is disputing any of this. But believing people can change doesn't mean ignoring what they do. House House are a very successful indie studio who have now made two enormous hit games back to back. They know half of the indie games scene, are celebrated the world over, and have sold a lot of games. All that is happening here is that they are possibly being disqualified from winning an award. Once. We need to put this in perspective a bit.

There are a lot of reasons why I think House House made a mistake doing what they did, but one I want to mention here is quite nicely summarised by this post from Simon Roth. Even before the thousands and thousands of layoffs, I knew dozens of freelancers who were desperate to find good work in the games industry. I spend a lot of my days scrabbling around finding little pots of money to fund freelance hires for research, or advising students how to survive in an industry that increasingly doesn't care about programmers. Unless I've misread their situation entirely, House House is not a studio that couldn't afford to hire a specialist for a few days to look into this problem. House House aren't laying off developers to replace them with Claude, and I'm sure they would be firmly against the mere thought of doing this, but we can also clearly see how their actions are part of the same problem. The Venn diagram of people defending this use of LLMs but who would've posted 'QA are gamedevs' in previous discourse cycles is basically a circle, I think, and this is frustrating - you can't just pay lip service to the idea that we're all in this together. It has real meaning for people who are out there watching contracts dissolve as everyone posts about how, well, AI for code is actually a lot more ethical, you see.

It's also pretty clear that House House were aware of exactly what they were doing and how it would be perceived, which to me suggests they don't even particularly agree with the defences of them being posted online. Yes, Valve do not require a public declaration of LLM use, but there are lots of important political stances that House House have taken publicly without being asked to do so. Releasing a statement precisely when there was direct public blow-up from IGF jurors taking a stand to me suggests that they do understand the situation quite well, which does make the lack of earlier disclosures feel a bit less ignoreable. In short, I think House House themselves probably have a more negative read on the situation than most of the people posting about it today. And I think that's cool because hopefully it means they'll figure out what they want to do differently in the future and come out of it feeling better about their process - even if that is "we want to use AI and we're going to tell everyone that".

And this kind of brings me to my other feeling on all this which is how this is being dissected online as some extreme artifact of cancel culture.

Not Caring About Anything Is Good, Actually

The worst takes I have seen on this matter attempt to frame it as an extreme position on the part of the jurors who released the statement, or unfavourably compare this issue with another important issue, or toss around phrases like 'purity test' or therapyspeak. I really feel like we've lost all sense of perspective on AI as an issue. The IGF has no AI policy. It explicitly is asking people to draw their own lines. This group of people drew one. That you disagree with where the line is doesn't make it somehow extremist or unreasonable. We're also talking about an awards show that some notable game designers have been asked to judge for. They are not going to raid House House's offices. Their accounts aren't getting frozen. We're not sending in the feds. In fact, one of the recurring issues with the IGF is that critically acclaimed games sweep the awards and then it doesn't do anything for their success. House House are already a wildly successful company, and they will continue to be in the future - even if they pivoted to using AI for 100% of their code.

It's particularly galling to me that House House are getting the smol bean treatment online when many of the signatories of the jury statement are freelancers who rely on the social network and goodwill with other developers to remain employed and pay rent (in some cases, doing so through coding). These are people who are giving up their anonymity to publicly stake part of their reputation, and by extension their income and survival in the industry, to simply explain how they are approaching their job for the benefit of the community. They could quite easily have done this in silence! This is how creative movements evolve and change. They are doing the hard work of actually addressing difficult shit out in the open, which is what everyone else has been trying to avoid.

Before I wrap up, I wanted to explain one additional thing. I agree with the jurors' statement, but there is one thing I think I differ on to most of them: I don't care about changing anyone's mind any more. I don't say this because I'm proud of it or because I think it's good, I say it because I'm sad and I wish it wasn't the case, but it's where I am. I am deeply tired and truly have no emotional energy for it, and the casual way in which people have normalised and accepted the worst shit imaginable has permanently broken parts of my brain. I've accepted that most people around me are going to increasingly find it okay to use AI, and it is going to creep more and more into their lives and work, and that includes the games industry. I have accepted all this and more. 2026 has seen a surge in normalisation that I simply could not have predicted. I don't think it will be forever, but it will last a long while.

I'm trying to channel the energy I have into supporting spaces where other ways of using, thinking about and making technology can exist. Not to the exclusion of all else, like you can dip in and out of PROCJAM if you want, but while you're in the space you are asked to follow the rules. I think these spaces are useful for lots of reasons, and I think they can be maintained and protected even in the presence of massive societal change everywhere else. I don't think this is the only answer, I just think it's a concrete and specific thing I can and want to do, and have the energy for, and I hope that it complements what other people are doing elsewhere. But often we hear the concern of "you won't change people's minds by being judgemental/hardline/etc" and, with the best will in the world, I am not interested in that any more.

AI uptake is expanding, even if the fate of the AI industry as a whole is unclear. Commentators like Ed Zitron have promised for years that everyone is going to realise they are being lied to soon, and the whole thing will just quietly go away overnight in an economic puff of smoke, but this is not going to happen. The problems that AI is bringing are going to be here for a long time, and any long-term community planning or organisation around this issue has to take that into account. I don't know if the IGF will be one of those spaces that resists AI work. It doesn't have to be, it won't be the end of the world if it's not. But I think it is the time to decide one way or the other, because the current state of affairs doesn't benefit anyone. And then we can move on, and different groups can either rebuild a new space, or redecorate the one they retained.

Thanks for reading. The Procedural Generation Jam takes place December 4th-13th this year. You can make a game, or a little procedural generator, dice tables for a TTRPG, you can make all kinds of stuff. Just leave Claude at home. Join up here.