Full agree. I don’t have the karma to unflag it, but I’m disappointed in whoever did. It’s *very* related to this forum and *very* topical to current events.
Yeah, tell [email protected] when it happens, they usually act on what you report (I already did for this post).
When it gets unflagged though, if it some hours old, it often has little chances to get back to the first page; this part of Hacker News is really flawed.
This is the part that keeps being frustrating. Everyone talks about AI Art and AI Code as though the art or code are either 100% ai-made or 100% human-made when they’re talking on these topics; but, for example, is it really AI art if someone uses AI to make the in-betweens in a hand-drawn game? Is it really AI art if they use context-aware background removal to clean up some mistakes on a scan or photograph they took?
> [...] is it really AI art if someone uses AI [...]
It would make clear communication even more difficult if we shifted the goalposts to the point that "using AI" does not constitute creating "AI art." Perhaps there is some conversation to be had about what constitutes "slop," or what counts as "vibecoding."
I hear that, and think that’s a useful discussion starting point. My biggest concern with that is that it’s like Prop 65. The most honest answer for nearly everything, though, is “this was made with the help of AI”. Even Google Searches aren’t immune anymore. Everyone has their own line; and even the “None” people, I think, don’t actually mean “None” by in large.
Interesting point about Prop 65... Part of my issue with AI as it exists in 2026 is the blitzkrieg approach taken by the big corporations, with regards to your comment about how even a simple Google search gets handed off to an LLM, whether you wanted it to be or not. The inescapability is intentional, and I think is being done (in part) to help drive Prop 65-like attitudes in the general public -- corps are training people to ultimately ignore AI usage declarations entirely, by making them borderline useless to anyone but the most extreme anti-AI absolutists (is that the right word?).
I can't get behind the defeatist approach of "well... the cat's out of the bag, not much we can do about it shrug." But I'm not sure exactly what approach I believe is warranted, or would be effective.
I think the other big problem is that, even if the tide of public opinion turns, there isn't really any way to verify a dev's claims that they did not use AI to generate some code for their video game, for example. That said, I'd still like the ability to participate in communities -- no matter how small -- made up of people who share my absolutist view, so I think a Prop 65-like warning would still benefit me personally.
More reasonable stances and recognition of consequences!
I don’t know how much is “being pushed on us to create complacency” compared to “because LLMs are just better at semantic search” (for Googling and doc search specifically); but cost reduction (chat bots for support) absolutely.
I am convinced that the aggressive branding of GenAI as AI *is* a long game to complacency-wash the new LLM techniques and all the data they’ve stolen.
As for myself on your last topic, I would love a sliding scale that a user could put themselves on from:
* Not even the tools I use were made with AI (ignoring operating system because that ship has fully sailed)
* The tools I use were made with AI, but I didn’t
* I used it to look up stuff / teach me
* I copy pasted a tiny tiny amount relative the whole project
* I used it to write scripts to ease busywork
* I collaborated with AI in code / used smart filters
* I prompted but was deeply integrated in the design process (agentic software dev with intact code review is here)
* this project used some fully AI assets in its final form
* this project used mostly AI assets in its final form.
But even for that, compliance is a real problem, blah.
* don’t want DLSS?
* don’t want your artists to use Photoshop’s newer features?
* don’t want your artists using Blender plugins that happen to
maybe have been made with AI?
* don’t want a game with hand-made clay models that are scanned in and
cleaned up?
* don’t want a game with hand-drawn animation that has AI-generated
in-betweens to push their 10fps drawing to 30?
* don’t want to play games on engines written with AI? (Unity and Unreal)
* don’t want someone to have used an LLM or a search engine
to learn something? (Google’s Fast Answer counts)
Because “None” is doing *a lot* more heavy lifting there that I don’t think you’re recognizing.
I think most people want art to remain art, and for a human hand to create the experience.
I'm not the person you responded to, but, yes to all of your above questions. This is a Nazi Bar situation. We must stop separating "AI" as a technology from the corporations that are pushing it (and pushing their way into governments). Right now, we absolutely do not have one without the other and our society is not equipped to separate the two quickly enough to prevent catastrophe (in my opinion).
Fair enough! That is an internally consistent stance for None, and that’s all I was asking for when people say specific things.
I may not agree with you; but you *are* internally consistent and that’s all I can ask for. My common belief is that people don’t actually believe “None” when they say none, in a similar way to “The only moral X is my X” or “but that’s different!” cases, and I try to put a mirror up to people’s declarations.
It’s unfortunate that most companies pushing it are pushing it from, at best, an amoral (not immoral, I hope) perspective, ignoring real fallout and negative consequences - or actively causing them in some cases - because I genuinely think that good has/can/will come out of transformer / generative systems, but there’s just so much collateral damage right now, and that could have been avoided.
I feel like if you’ve thought that far through it, you know which companies to resist and argue against, and which are collateral / easy-to-bully, but bad targets.
Putting words in your mouth, e.g. House House / Big Walk:
* should they be removed from the award? Yes. They used AI, even a
little bit.
* should they be Steam Review attacked and have their name
plastered everywhere? No. They didn’t create the environment
that they operated in, but they did break the rules of the contest
and admitted it.
> Because “None” is doing a lot more heavy lifting there that I don’t think you’re recognizing.
Your examples are all nothingburgers to an anti-LLM guy like me. It's easy: Declare your slopbot use transparently (authorship), and I'll act accordingly and will pay no more attention to your output/product, if necessary (i. e. in virtually all cases). Don't declare it and it's also simple: I, or someone else, asks. Then it's either deception or omission. Which brings us back to acting accordingly. :)
Some of your questions roll into pathetic triviality/hairsplitting territory (e. g. somebody else's search engine use), though.
You’re asking people to declare *all* of their usage. Therefore, using a search engine, using Unity, Photoshop, Unreal, Corridor Key are all appropriate declarations for a person who is absolute-honesty minded, which is a *very* common state for programmers. Think of how often you see on this very forum people trying to Rules Lawyer the law when it usually doesn’t actually work that way.
Under your position, how do you protect against a Prop 65 world (or a world where everyone hides their AI usage because all of those above *do* count). They counted enough for House House to be functionally disqualified because they asked AI to help with a bug - and they *didn’t even use the code it gave*.
> You’re asking people to declare all of their usage.
Sure. Declare it all.
> Under your position, how do you protect against a [...] a world where everyone hides their AI usage because all of those above do count.
A world of nothing but drooling deceivers? You might as well bring on the apocalypse. I'm sure somebody's written a book about that.
> [...] House House to be functionally disqualified because they asked AI to help with a bug - and they didn’t even use the code it gave.
Never heard of 'em. Poor buggers. Well, you live and learn as my ma used to say!
EDIT: Checked up on it a bit. Wiki sez [1] that, to quote, "'LLMs were used sparingly by House House to help work through difficult technical challenges and bugs', that they 'estimate that well under 0.1% of Big Walk's code had anything to do with this usage', and that 'generative AI was not used in any other context when making the game'."
That sounds a bit different than what you described. ;)
> Games being difficult to produce was an important feature of the artform. Not a caveat.
Only in that it meant parasitic- and slop- groups couldn’t harm makers (parasitic) nor players’ experiences (slop).
For everything else, a high barrier to entry for a video game is a detriment and does not reflect the game-making industry at large which does *not* have that barrier, at all.
Board games, card games and pen-and-paper games are notoriously *VERY* easy to make for yourself, as you literally only need some of cardboard, cards, pens and paper.
I’m decidedly against this “everyone has to go to the exact same school of hard knocks that the past went through” when it comes to gaining a skill or trying something new. Creating a game is still quite hard when it’s someone actually trying to make their vision real.
Working backwards from my understanding that one of TFA's assertions is that AI is problematic when it is used to take a human's job: could frame gen DLSS affect human jobs?
Without DLSS, the game's performance would presumably suffer, which could necessitate that more human workers are hired to improve performance somehow. Therefore, I think the logically consistent view is: frame gen DLSS is unethical.
I'm replying to myself to expand on a counterpoint that came to mind: if frame gen DLSS is unethical because it allows companies to avoid hiring humans to improve performance, are technological advances that allow higher clock speeds on GPUs, faster memory, higher memory bandwidth, additional cores, etc., also unethical on the basis that we can replace human effort with technology?
I think the logically consistent view must be that, yes, technological advancement that displaces human workers is unethical. Full stop. Unless there is an argument to be made that certain negative externalities are tangential enough that they can be ethically ignored, or our society has developed systems that lessen the harm of this displacement to an acceptable degree. But then what would "an acceptable degree" be? Filing for unemployment?
Hm … isn’t this more of an argument toward a better social safety net?
Going full StarTrek, Replicators put lots of people out of a job; but their net improvement in a society with a very robust safety net is very, very visible.
To a smaller extent, one could argue that things like 3d printers, even, reduce reliance on firms that specialize in small-scale manufacturing
Uncanny Valley does not accurately describe generative green screen removal or background fill, both of which are GenAI and regularly ignored in this conversation while “ban all AI” is the only nuance many people are given.
I would argue signal to noise will greatly decrease, but we will have amazing games we never had before because artists and non-programmers will be able to make their vision in a world where the barrier to enter was too high.
But there will be a LOT of slop we’ll all need tools to filter.
If it's anything like the 'democratization' of music publishing, discoverability will indeed become the real problem. Most games will languish unplayed (probably deservedly), just as the vast majority of tracks from providers like CD Baby are never listened to.
I’ve been following the Big Walk and VRChat fallouts around AI Usage and my response and thoughts stay mostly the same:
* how much is too much usage?
* how much do you need to disclose?
I think this article sums up a cohesive and defensible stance, mostly (the environment parts are shaky, imo), especially around overwork and cognitive strain for devs.
I continue to be frustrated about how they never mention how useful it is for learning when you use it as a side tutor, or how useful specialized tools inside it can be, like Corridor Key 2, a green screen removal tech; and the blanket advise to “show disapproval of all use” is a path toward the witch hunts we’ve seen already.
But they definitely make strong points about overwork, increased velocity, burn out, etc. Major problems, especially in the gaming industry, and I hope this can be another angle to finally give mercy to them, though I have my doubts.
All in all, I think it’s good advise, expect for the disdain stance near the end. Hating on coworkers or makers rather than the systems and power structures that are making things worse only hurts the person doing their best, imo.
Edit: to the downvoters, why? And more importantly, if your comment is “well of course we don’t mean *that* usage” then please, define your terms because even this argument doesn’t describe whether they’re meaning educational, facilitative or full vibe.
Is there already AI-generated/AI-assisted code within Unity, Unreal Engine, Photoshop, Maya, etc?
Because if there is, it's already just not practical to attain the level of AI-free 'purity' that some are demanding.
(Maybe instead of focusing on the developers experimenting with new tools, focus on the big AI companies and the masses of copyrighted content used as training data?)
There is AI in at least the first 3; and Photoshop especially has AI tools, so it’s not just using a tool that used AI in its production, it’s using GenAI to make your own assets.
> (Maybe instead of focusing on the developers experimenting with new tools, focus on the big AI companies and the masses of copyrighted content used as training data?)
How do you suggest an individual act to achieve this, seems like divesting from AI and pressuring other people also do so is a good step in that direction.
A major part of the problem is that the only people that will feel this pain are the small companies or small teams. See the absolute dog-pile that House House, the company that made Big Walk, has been experiencing.
It basically taught every developer to hide their use of AI if they’re going to have that kind of backlash.
For what it’s worth, there are authoritarians on most any binary - people who would demand compliance through considerable use of force rather than goodwill or trend-design that makes “the desired path” the easiest one.
reply