PC Gaming

Whilst for my project AI Gen was only ever an idea for a nice to have which is not important for game-play, I'm pretty sure that there will be projects out there being done by tiny Indies which aren't financially feasible without AI Gen because those operations are not well funded and can't afford to pay for lots of manpower.

In game-making, generation tools (not necessarily AI) even the field between Indies and AAA game makers (which is why so many Indie titles in this latest blossoming of Indie Game-Making have procedurally generated worlds/levels whilst the AAA titles almost invariably have massive hand-crafted worlds/levels) but until AI Gen the unassailable advantage in favor of the AAA makers was in the finishing touches - for example, it has long been possible to use procedural voice generation, it just doesn't sound as good as the stuff done with ML (unless you're making a game about robots were a robotic voice does sound great) - since one can only go so far with procedural generation so in more real-world-related domains (voice being a great example) procedural generation is usually shy of "good enough" whilst both AI Gen and professional human crafted content is beyond it even if the former is IMHO generally not as good as the latter.

In gatekeeping a certain level of quality to only things that can be done by those who can afford to hire large teams, because you refuse to accept games made with the kind of tools that most benefit the smaller game makers, you're basically supporting what's best for the bigger companies, unless the only kind of games you buy are "text-only dialog and limited art assets" games made by Indies with small budgets (in which case I'll take my hat off to you for being Principled in a consistent way) and not the more glitzy stuff that only bigger operations can afford to make without AI Gen.

Merely being against the kind of tools that most benefit small operations and then turning around and mostly buying the work from the most massive of operations because it has a better quality (since they have the economies of scale and revenues to afford real human craftsmanship) wouldn't actually be a consistent principled stand IMHO.

In the game making world, gatekeeping AI Gen use outright "just because" is a great way to keep the playing field tilted in favor of the likes of EA.

The ethics based on Intellectual Property? Quality, sure, but ethics?

Full disclosure: I'm a geek from the days of newsgroups and Geocities. I watched the rise and fall of things like Napster. And I watched IP-law get more and more restrictive. But what is "intellectual property" really? You're effectively taking an idea and saying "this is mine, I made this first, therefore I own it".

Around 1996, when I was 12, I thought it'd be really cool to have a small laptop that laid flat and you could hold in your hands. The designs I drew up VERY closely resembled a Blackberry. Blackberry came out a few years later. If I had filed the right paperwork, at 12, should I be able to stop them? I sincerely doubt they were spying on the drawings I made on the back of my homework. Should you get to stifle innovation just because you had the first brainfart? I don't think so.

But okay, let's say you're only thinking about artistic works. Again, you're gonna have repetition. This came out in 1995. This came out in 2008.

So what's the issue with AI; it was trained on "copyrighted" material? K, well so were you. Are folks upset because creators didn't get paid every time an AI reviewed their copyrighted works? Well, are they similarly upset about folks who check a book or movie out of the library? Not so much...because that's normalized (though would NEVER go over in today's hyper-corporate nonsense world). Okay, so are folks upset that generative works can resemble the style or "essence" of the original work? Lol, see the Jill Sobule/Katy Perry comparison above, also consider "Fair Use" and the likely transformative nature involved as well.

This isn't an "ethics" issue...it's an issue of disrupting existing channels for corporate power within a world sliding more and more into a dystopia of corporate fascism.

I think you have to be clever with the usage of gen AI to get non-boring things, and just use it as one or multiple elements in a larger pipeline/computation graph. This is my intuition and not battle-tested.

Disagree. They can be connected to actual game mechanics. For instance, it's quite easy to ask an LLM to output something in json format:

{
  "name": "The Master of Evil",
   "hitpoints": 205,
   "class": "vampire,"
}

and so on. You might object that it could make mistakes here. Suppose the detectable error rate is 10% (I actually think it's lower from what I've played around with.) Rerunning it in the case of a such an error (e.g. malformed json, invalid class name, hit points exceeds bounds, etc.) reduces to 1%, then 0.1% etc., and in the end there can be a non-AI fallback just for certainty. Admittedly, the errors are not i.i.d., but still it should be pretty low. Many traditional procgen techniques, such as map generation, also use rejection sampling in this way, with even larger rejection rates than 10%.