The IFTF’s lawyer likely said something along the lines of “it’s unlikely but not impossible, but if it happens it WILL obliterate your organization so it’s best to take precautions”. I don’t really need to know the actual specifics since I’m not a lawyer and couldn’t do anything with them anyway, but I’m happy to assume the IFTF paid for good legal advice? At the time this was all completely untested because the law came into effect very shortly before the comp started, so I don’t think it’s quite fair to point at everything that has or hasn’t happened since then as evidence that anything should have been done differently.
And similarly, we don’t know these kinds of policies are necessary. Unlike the existential threat of the UK age restrictions, more blowups over AI are relatively easy to deal with – forum moderators have a lot of leeway to mute, delete, and ban anyone who steps out of line, and that includes anyone organizing a witch hunt offsite as there’s already rules about that (not that we need them, as Draconis has made clear - if someone invents a new kind of bad behavior that’s not already against the rules they have the power to deal with it accordingly). The Comp already has rules for authors about disparaging competing entries as well, although they’ve been relaxed in recent years, but anyone publicly stirring up drama about a competitor already has grounds to be punished. And besides, it’s seven months until the Comp! They could have a plan in the works already and just not be ready to talk about it publicly.
To be blunt, it doesn’t matter if we on the forum were getting anywhere – as I believe was mentioned earlier, the IFTF does not take action on anything posted on the forums so you’re better off emailing them with your concerns and suggestions like Dan did. I’m not sure there’s any point further debating this publicly in a space where none of the decision-makers will actually see it.
Not all of them are—but Google’s images include an invisible SynthID, which doesn’t go away after a screenshot nor a crop and can be detected by Gemini.
To me it feels like having a potluck group where everyone cooks together, but some people start a trend of using and bringing dumpster-dived food.
Dumpster-dived food can be perfectly safe; there’s a lot of waste out there. So you can’t really tell if it’s dumpster food or store-bought food, if they are careful.
But instead of being careful, you get six new people that crash the potluck holding dented, bursting cans and packaged chicken that’s green and bloated and rotten blueberries, excited to present their contributions. One person just brings a recipe for food made from dumpster-dived ingredients (the receipt is dumpster-dived too).
You could just say ‘I’m not going to eat any food that looks bad, but people are free to bring it.’ And that’s totally fair. You can say, ‘Some people are really good at hiding their dumpster-dived food, and some people’s home cooking is as bad as rotten food, so we shouldn’t discriminate.’ That’s also fair. But at least posting a sign saying ‘no dumpster dived food’ would keep out the worst of the worst.
Anyway, that’s the end of my analogy. And if you think, ‘but dumpster-dived food can be safe, and is an ethical way to deal with consumerism and to stick it to companies, and some people have no other option’, all of that can be true for sure, just like there are arguments for AI use; I don’t think either issue is black and white, and there are some uses of AI that I think are actually good, especially with image transcription and refining language for non-native speakers. I just don’t want mold at my barbeque!
I also hope I’m not giving too dictatorial an impression of the forum management! Actual issues involving actual forum members are few and far between; most of our job is looking at first posts with links or images in them and saying “yep this is IF” or “nope this is spam”, or deleting spam that got past the filter. (Lots of emails to the support address.)
That, and splitting topics on request.
But that’s also why I don’t expect many problems with the new policy. This community is tiny compared to Choice of Games, and the people who want to maliciously sneak LLM-generated games into IFComp are (I presume) a tiny fraction of that. I would be shocked if the organizers had to deal with more than two or three instances per year.
(Which may be way off-base. I do forum management, I’ve never been involved in IFComp organization. The organizers have said there’s been major vote manipulation in the past, which I also wouldn’t have expected. So take everything I say with a grain of salt compared to the actual organizers’ words.)
If a policy is explicit, it’s easier to enforce. Everyone understands why the decision was made and people have less leeway to argue or make excuses. What I meant about loopholes is that, if, for example, it’s decided that X amount of suspicion warrants a disqualification, we could just tell people that. They’ll figure it out through observation sooner or later anyway and at least by announcing it there’s clarity. Decisions won’t seem arbitrary and people can have some degree of confidence they won’t be wrongly targeted.
I’m not sure anyone in the IFComp realm would bother just because even people that have tried to use GenAI for a story have been pretty open about it and would completely respect the rules.
The consequences of attempting to circumvent the rules would far outweigh the effort.
We maybe had some trollish people back in the 90’s, but I’m not sure that behavior exists anymore.