There appears to be a lot of calls for nuance here.

This is what people are mad about, taken from the front page of nextfest: https://store.steampowered.com/app/5106420/Leverage_Real_Est... https://store.steampowered.com/app/5155900/Ages_TD_Demo/?cur... https://store.steampowered.com/app/5200920/BATTLE_PARTS__POW... https://store.steampowered.com/app/4609270/Naval_Supremacy/?...

These are, as a whole, poorly-balanced, poorly-implemented, and in some cases poorly-conceived. The art is terrible, but the code is also terrible, and the games are not fun. Not only are we lacking authorial intent, we’re lacking any amount of basic quality. Everything from the code to the store page is AI-generated, and it makes it not only be unfun, it robs you of the usual markers of a low-effort cash-grab asset flip (eg, Stock Car Store Simulator)

And there are a lot this year

What Steam users want is a way to filter out crap but crap existed well before AI (trust me, I've spent a _lot_ of time scrolling through games on Steam both for "Fests" and just in general). And unfortunately, there isn't an easy way to filter "crap" out. There as good games that use AI, there are bad games that use AI, there are good games that don't use AI, there are bad games that don't use AI. I don't think "AI" is a good filter.

Thankfully Steam has easy refunds if you don't like a game or it doesn't live up to it's marketing page.

I don’t have solutions but tagging AI is definitely not easy. A lot of art is expressed via code: particle effects, menu layouts, iconography, animation keyframes. AI having written any of the code would get so many games tagged as AI that the label would lose all meaning. More recently, Opus 5.5 is clearly a model that started to develop visual taste and can draw images with code, as opposed to outputting images as tokens or via diffusion. And these capability gains are still accelerating.

There’s a wide spectrum of AI art between particle effects and a character drawn with brushstrokes via code or computer use and it’s hard to decide what should be labelled and how the labels can be enforced.

How would no-AI filter be implemented by Steam at all?

Do people expect Steam to proctor the entire development process?

Issue here is not AI use, but very low amount of creative human effort that went into making some games.

There are three existing realistic solutions for this problem:

- raise SteamWorks deposit of $100 for launching a new game on Steam (to reflect lower cost of making games with coding agents)

- reviews by users and pro game critics

- users can request refunds under 2 hours of playtime

The video game market was already flooded before LLMs. Now, it's much much worse. I'm not sure that "AI" vs. "no AI" is really the right filter, but I do understand the need to filter out the chaff. There's too much content, and too little time. Even for something that you love, there's more out there than can possibly be managed or enjoyed. This problem will get worse.

Isn't it a gradient? I'd probably rank it like this.

AI Used in back-office administration, internal documentation & presentations

AI Used in research

AI Used for generating placeholders

AI Used for Upscaling (DLSS)

AI Used for translations of UI.

AI Used in the final product (assets incl. textures, 3D, sound and voice)

AI Used in final product (code)

To say nothing of the other disciplines of AI like machine vision or voice recognition or dynamically building a story tree or character response.

People need to understand that AI is dominating the corporate world; all areas, not just software development, take advantage of AI. Denying AI in the corporate world means losing competitiveness. You can't fight against the modernization of tools. And game development is no different; it will accelerate and improve games, but it will also cause a flood of flop games. Game design, good ideas, creativity, common sense, care/attention, make a good game. AI only accelerates, for better or for worse.

AI or no AI is the wrong metric, of course. I imagine the way to solve this usefully probably goes through something that looks like Reddit. More "meta-moderation."

The problem is "a number of submissions we weren't prepared for," not "AI."

Steam has had problems with asset flippers long before “AI” hit the mainstream, but due the fact that even those “human slop” took time to make, you didn’t see them as often.

Sure, we at HN are holier than the rest and we know how to “bend the powers of AI for good”, but what Steam players see is a sea of mass produced AI slop games which barely work at all.

And yes, there is that one guy with good storytelling skills using AI to fill the knowledge gap in other areas make a game with some interesting ideas, but even him is getting overshadowed by the sea of slop. (Still a point could be made here of why. Why do everything alone? Why not take a chance with an aspiring VA recording lines from their phone? Pretty sure some people would work for free to get their first chance to get featured in anything)

We know it makes absolutely no sense to slap a binary “AI” sticker in a game just because one dev asked Claude to help them fix some scripting bug or whatever but maybe, just maybe, the best solution in that case would be outright lying. We know who that tag is for, something something read the spirit of the law, not what was literally written.

Right now I am glad I am not a Valve employee because they will have to figure out some scaleable solution for the (lack of) QA problem they have been ignoring for a long time now.

They really want a "no dogshit" filter, I imagine. So a filter for actual effort and care.

Slop is inversely proportional to time invested.

But yeah, already a year ago you would see on Steam, twenty almost identical games released on the same day.

--

As a counterpoint, I spent several years working on a game that's absolute shit. (And I can name similar examples but I won't!) Meanwhile I've made much better games with AI. (Or rather, AI made them...)

Quality is decoupled from time invested. That was already true before AI. But time is still probably the best proxy we have.

Maybe bringing back Greenlight would be good. Nobody would Greenlight 10 identical puzzle games uploaded on the same day. But they would accept high quality games, regardless of how they were made.

This is your "abundance".

People are proud to use AI but don’t want the AI label. Seems sus.

"Trust (customers) but verify (games)" works but at the cost of Steam/Valve having to implement either human verifiers or AI verifiers.

This will cost with low ROI.

If Steam is indeed customer focused, they might but doubtful.

I don’t know how these filters don’t inevitably or immediately become a hellscape of fighting over definitions and enforcement…

Ironically what I think people really want here is a better recommendation system, one that almost certainly would need ML algorithms (or 'AI' as we call it nowadays) to implement.

So Valve will need to train on descriptions of games and perhaps demand more out of developers who want to sell games on their platform.

All notable app stores face this already or will face it soon. Maybe the apt GUI for your favourite debian derivative won't get too flooded...

AI doesn't make games good or bad. People make games good and bad.

The expectation of quality is slowly being chipped away. And it appears the HN community is prepared to “nuance” that expectation out of existence.

I feel for me to even have this conversation we need some gradient we can point to and say this is clearly all AI and this is clearly minimally assisted and everyone is along this gradient and past some arbitrary point its no longer aggregable to me.

I use claude in my games(Currently a hobby) I draw a character on paper, claude models the proportions symmetrically and then I go back in and slowly build up the model. I wouldn't have a problem with artists doing this at all, in fact I don't know of a good reason not to. However if they generated a picture, sent that to claude, took whatever claude generated and threw it into the game then its going to look like garbage and cause they didn't start by throwing new information into the system it must start from training data so it will look the same as all the other slop.

That first step of starting with a piece of paper or some human seed is huge, the more detailed that step the less I care about the rest of the process.

AI is a tool nowadays, cant ban it.

plenty of games made with AI that are fun.

Nobody is going to mark their game as "Made with AI" unless customer sentiment shifts or there is a very strong incentive not to lie. Right now releasing a game with a ai label is just net negative. Meaning the filter would be utterly useless even if it was introduced.

It might be time to bring back steam greenlight. Some games are made significantly better with AI as the developers are very bad at programming, but they're damn good artists and have a vision. Not everyone has the money to hire capable people to bring their ideas to life.

Introducing a better review system where you could rate how fun, asthetically pleasing and whatnot a game is would likely help with sorting out the sloppy mass produced media. I think steam could take one step further and build a classification system based on required-to-submit 1 hour long of in-game footage. The effort needed for sloppy games is far lower than recording 1 hour of gameplay in 90% of the slop and I believe that would at least somewhat make people reconsider knowing that they can't hide behind a fancy trailer or cool visuals.

Steam had a problem with trash even before "AI". It allowed rampant asset flips that required little to no technical expertise and certainly no effort. "AI" tags make no sense because you cannot separate supposedly "vibe-coded" slop from any other game made by anyone from normal indie developers to big studios (and they are using the same tools supposed vibe-coders use). Disclosures are also useless, because the entire development process cannot be meaningfully audited and it becomes a gamble people will make to try and not get filtered and such. Valve needs better curation, but it did not do anything about it before and it certainly will not do so now, evidently it's profitable this way.

IMO, We really should be centering these discussions around the lines that separate Remix versus Ripoff. I want to live in an age of less effortful remixes that push the culture forward, but we are mostly getting ripoffs that slopify the culture at the moment. Shun the ripoffs, praise and boost the remixes, same as it ever was.

maybe judge games by if they are fun or not

I'm participating with my game in this Next Fest and I don't really think this is much of a problem. The amount of games presented doesn't seem significantly higher than last Next Fest. Games that are good get boosted after 2 days (first 2 days everyone gets featured randomly) so low quality AI slop won't get much of attention. And the way they check if a game is popular is wishlist count and player/install count as far as I understand. I don't think low quality slop will get much of boost, and if it does it probably is deserved unless the wishlist count/player count gets botted somehow, but I haven't really seen this happen but who knows. For most people I think it's best to wait until after the second day to see which demo's are worth checking out, there won't be much AI slop there.

Maybe they should just focus on "is it fun".

Gamers and game journalists seem to be unaware that most software engineers are using AI now, likely including programmers who work at prestigious AAA games. But they only see the vibecoded slop, and assume that AI=slop, when in reality ~all games use AI coding now, both slop and non-slop.

“No AI” is too simplistic.

Does that mean no AI assistance at all? How is that possible to detect for if someone uses it for code generation and compiles it to a binary?

Do they mean no AI artwork? Maybe easier to detect, but mostly based on vibes still.

Do they mean no AI features? But why, wouldn’t an adventure game where you can talk naturally to NPCs be kind of cool?

Madness.

How about a filter for “No souless shitty copycat games that are pure cash grabs”. I’m sure Jev can help detect these at scale.

I think the number of submissions is fine. It's easy to filter by the type of game you're looking for. If you just want to doom scroll, yeah you're going to have a bad time.

One of the things that captivates many people who play games is imagining building their own.

It's understood that there's a massive new flood of games from new game creators that are using a higher level abstraction like AI to build games.

At a smaller level, it can sound like the differences between folks that used to build games using assembler only or a higher level programming language, which built to the point until there were higher and higher level tools.

Speaking of low effort, "ban Z" does not constitute a thoughtful and intelligent solution to "Z created a problem."

Z=AI in this case, but lazy solutions are as old as time.