Requesting an AI statement from IFTF

Competitions IFComp
1

Before IF Comp, I mentioned that I wanted IFTF and/or IF Comp organizers to publish their full stance on the use and disclosure of generative AI.

At that time, I wrote:

I’d like IF Comp to fully describe its stance on AI use outright rather than just attributing it to direct and indirect pressure from the community.

The news post suggests the [AI] disclosure rule is just a convenience for audiences (“This information will be shared with judges, so they can make an informed choice during the voting period.”) However, based on the current and past threads … I’m under the impression that concerns around reputation and public accusations influenced the decision.

On top of that, if IF Comp agrees with broad ethical and ideological concerns (risk of theft/plagiarism, discriminatory content, devaluation of human labor), it should say that outright. Or, if it is neutral on AI use, and it’s meeting community demand simply because it serves the community, it should say that outright.

I was reasonably happy with the way IF Comp turned out this year, but I’m not happy with the initial statement at all, as it gives no indication as to how things might develop in the future.

I’ve mentioned that I’m against mandatory AI disclosures and most other types of mandatory disclosure. However, this request isn’t an attempt to fight the current rules — just an attempt to clarify motivations as much as possible and hopefully prevent rule creep.

Anyway, is a full statement something that IFTF or others involved in IF Comp would consider doing?

3 Likes
2

I have to agree. Artists should be able to present their work as they see fit and not subject to red tape.

3

That’s a silly way to describe the behavior of “We think inflexible and forever-unchanging rules will make a space rot into an unmanageable mess under the overwhelming pressures of passing time, trolls, and armchair lawyers, so we would like to remain dynamic to prioritize the health and cohesion of the community, so we can continue to exist for the long term”.

6 Likes
4

Also, in case my original reply gets marked by a mod (fair reason to, probably), here’s an alternative:

That’s a novel way to describe a target which defends itself.

1 Like
5

Some of the relevant people (e.g. the organizer for IFComp) aren’t often on this forum, I don’t think. You might have better luck reaching out directly.

4 Likes
6

The IF Comp organizers justifying the rule with a single sentence that attempts to shift their motivations to the community justifies a fear of rule creep, IMO.

I don’t think that a published statement would prevent IFTF and the IF Comp organizers from making decisions on their own terms, and I realize they have the power to choose the rules.

However, if they ever did decide to discuss further changes with the public, it would give everyone a starting point to work from.

7

I feel like there is a fundamental issue behind having a fear of rule creep at all. That’s what I was saying.

6 Likes
8

I dislike to discuss AI, but for me the core issue is elsewhere, namely code generation & optimisation.

Sure, the current and predictable future IF compilers are much simpler than GCC, but the latter already use heuristics and pattern matching in these crucial phases. and the open source community isn’t much far behind the closed-one (or, if one prefer, corporate) development of AI (and the current global worries on AI & corporate abuses of AI will motivate the OS community in quickly closing the gap)
(yes, I’m writing from the political scientist side of my peculiar polymath…)

Perhaps came from my EU mindset, but I think that innovative but potentially distruptive technology is best handled by reining than resisting, that is, regulating (the core concept of EU mixed economy) but personally I dislike “preemptive regulating” and I feel that the spate of AI vs. IF debate (which I prefer to generally watch from the sidelines, if not even ignore) is turning in this direction, as the debate involves the “regulatory organs” of the IF community (not only IFTF, but also the competition rules)

IF are based on text and text is relatively easily machine-handled, provided that nonsense like EBCDIC is kept out of the picture, so, outside actual creative text, AI can actually have its usefulness, more so if used at the simplest level (Personally, I will more than welcome an expert system capable of pointing to bad english, and I note pointing, not correcting.)

(Little digression: I think that the community’s competence is capable of coding an IF-specialised linter, capable, e.g. of warning about non-fixed door (the classical major mistake…)

so, my stance is that AI in an IF environment, that is, an environment in which the creativity is based on the same building block of the coding (and intermingled together; I remain convinced that the “zarf license” is the best license for IF work, that is, characters (and digits) put together in a meaningful combination (Inform 7 & 10 is a major case in point) should be regulated in a way consistent with this peculiarity; the consistency lying in balancing between the “narrative and crossword”.

an expert system which in noticing, say, the word “door” in the narrative, raise the depth of analysis on portability attribuites in the code around IS a major linting tool in the IF context, and I hope that everyone agrees on this.
Conversely, if code (and I stress code) generation & optimisation has AI elements (heuristics, pattern matching) is welcome, esp. in the often constrained environment of IF VMs (the lease of life to .z3 VM from the unused code omission IS a major case in point, at least IMO)

so, I think we need more intelligent toolset for the coding side of IF creativity, than debating on regulating AI abuse; the time saved in coding and debugging is time gained for the creativity (side point, fooling around IN ITALIAN with Twine has pointed at much more time spent chiseling the text, and again, this in Italian.)

but for English-speaking people, I say only that Rule 16 applies, esp. in the light that AI tools are mainly from the western side of Rule 16…

Hope of having explained well my viewpoint, and

Best regards from Italy,
dott. Piergiorgio.

1 Like
9

I’m somewhat confused what would actually be accomplished by making the entire IFTF take a pro-AI or anti-AI stance—aside from driving off everyone who has the opposite stance, which they seem to be specifically trying to avoid.

IFComp also requires you to specify whether your game is parser, choice, or both/neither, and that rule came about from the schism around the introduction of Twine. And to my eyes, at least, that seems to have achieved its goal. Now people who don’t enjoy choice-based games can simply not play them, and vice versa for parsers.

It’s not like removing the warning will make the anti-AI crowd suddenly enjoy AI-generated IFComp entries. So why not just let them skip it, instead of playing it, disliking it, and giving it a low rating?

12 Likes
10

Reading the discussion above, it’s really not “the entire IFTF” at issue. This is a question being asked of the IFComp organizers.

The news post (from August) says

As always, we will welcome feedback after this year’s competition in our annual post-competition survey

So I hope you did that. Or you can send them email at their contact address and let them know what you think.

I’ve been around through a bunch of IFComp rule changes, and they’ve all been done in the spirit of “Well, let’s see how it goes this year, look at the feedback, and then decide whether to make adjustments next year.”

It’s a bit early to expect the 2025 rules announcements, though. The last comp is barely over and the organizers do like to take a few months off in between…

12 Likes
11

I agree. Have you made this request directly to them? If not, this sounds a lot like you’re whistling in the town square trying to make people take sides.

Here’s the contact info if you aren’t aware:

Contact us

They may not get back to you quickly, but I know they welcome feedback that helps them make decisions.

Any author who doesn’t agree with contest rules can make the decision not to participate.

If you’ve not reached out to them and are just calling for debate here, then this is just fodder for yet another AI pro/anti debate thread.

8 Likes
12

@Draconis IFComp also requires you to specify whether your game is parser, choice, or both/neither, and that rule came about from the schism around the introduction of Twine.

As you noted, there’s a third category (both/neither), and several entries this year had no category listed, meaning it’s not mandatory.

Of course, the choice/parser schism has died down, but authors still might not want to be ‘judged by their cover’ even if it would be easy to apply a choice/parser label.

@zarf Reading the discussion above, it’s really not “the entire IFTF” at issue. This is a question being asked of the IFComp organizers.

Sorry for being vague on that. It’s unclear at what level the rule originated … I assume it was among the IF Comp organizers, but I didn’t want to assume that. I know that IFTF also runs other sites/events, and I don’t know whether rules are being introduced there.

@zarf I’m somewhat confused what would actually be accomplished by making the entire IFTF take a pro-AI or anti-AI stance—aside from driving off everyone who has the opposite stance, which they seem to be specifically trying to avoid.

It doesn’t necessarily need to be pro- or anti-AI, but it should be as thorough as possible based on the facts of the situation. I’d prefer something closer to Steam’s statement than Itch.io’s statement (ie. the style of justification, not the particulars of the disclosure rules).

Hopefully taking a stance would spur the committee and/or IFTF into more decision making. Even if one group of people is unhappy with the result, I think that all sides would agree that AI policies are something big enough to merit a broad long-term plan rather than year-by-year and case-by-case decisions. (To the extent that’s possible with the time and resources available.)

@HanonO If not, this sounds a lot like you’re whistling in the town square trying to make people take sides.

I did want to comment again publicly, but I realize now that I wasn’t just overlooked in the earlier thread — but rather that some of the organizers are absent from the forums. I’ll contact them privately. Thanks.

Mods can lock this thread.

1 Like
13

Concur and second Parjeter’s motion of locking this thread.

Best regards from Italy,
dott. Piergiorgio.