191
Steam updates AI disclosure form to specify that it's focused on AI-generated content that is 'consumed by players,' not efficiency tools used behind the scenes (www.pcgamer.com)
submitted 9 months ago by to c/pcgaming@lemmy.ca
82 comments fedilink hide all child comments
you are viewing a single comment's thread
view the rest of the comments
[–] 48 points 9 months ago (45 children)

Big difference but I would argue both require disclosure because I will opt out of any of it. Add it to the long list of bullshit in the gaming industry I will not condone with my money.

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 45 child comments
    •
  • [–] 36 points 9 months ago (2 children)

    The problem is that it's unenforcable. I bet that's one of the reasons valve is rephrasing.

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 2 child comments
    •
  • [–] 15 points 9 months ago

    Even pure AI art is unenforceable unfortunately. Like any form of cheating, some will be amateurish and obvious. But others will be sophisticated, skilled, and will simply blend into a gray area where you can't easily define a line.

    How much "AI tool assistance" does it take before it's called "AI generated content"? It's totally arbitrary, and in many cases it's going to be completely unenforceable.

    That doesn't mean it has no value, but it does mean it's not a silver bullet and no amount of tweaking is going to make it one. We can quickly use it to take out the obvious slop, the well-crafted examples will pass beneath anyone's notice, and when examples fall into the gray area we'll all bounce around inside with arguments about who we believe and how much is normal and acceptable until we eventually reach an arbitrary, per-game consensus, or maybe adjust the "rules" a little to accommodate them, but nothing really changes, we'll probably be arguing about whether games contain "too much AI" for decades now and there will never be a clear solution or answer.

  • permalink
  • fedilink
  • source
  • parent
  • [–] 25 points 9 months ago (6 children)

    https://www.reuters.com/business/nearly-90-videogame-developers-use-ai-agents-google-study-shows-2025-08-18/

    Good luck finding a dev that doesn't want to use/ isn't forced to use / doesn't lie about using AI tools.

    Ar this point if we're to shun all AI tools we might just give up the hobby.

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 6 child comments
    •
  • [–] 18 points 9 months ago

    While on principle I don't care about people using llms to refactor code in my games, I still think that the AI is inevitable narrative is a bit jarring and that study in particular has a huge conflict of interest issue.

    👍 1
  • permalink
  • fedilink
  • source
  • parent
  • [–] 4 points 9 months ago (14 children)

    You'll need to opt out of pretty much anything digital than because almost every business is has employees using AI is some form or fashion since it's shoved down everyone's throats so hard.

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 14 child comments
    •
  • [–] 6 points 9 months ago* (13 children)

    Anything that I find that's digital and uses AI, I do opt out of, thank you.

    I called an HVAC company several weeks ago and they had an AI agent answer the phone. I hung up and called someone else. No problem.

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 13 child comments
    •
  • [–] 5 points 9 months ago (12 children)

    You better stop using lemmy or your lemmy client then.

    Odds are astronomically high that they've used AI at some point on its development.

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 12 child comments
    •
  • [–] -1 points 9 months ago (11 children)

    I don't use Lemmy, I use PieFed, and it never used any AI. Thanks for playing though.

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 11 child comments
    •
  • [–] 8 points 9 months ago* (last edited 9 months ago) (10 children)

    You really don’t understand.

    They’ve used AI too.

    Developers are technology forward people. They try new things out. Maybe there’s some unit tests in it, or maybe they asked ChatGPT some questions to see how it responded when they had a problem vs a regular google search. Maybe they did a google search and the answer was in the AI response at the top.

    Their IDE probably includes AI tooling that automatically functions like code completion / suggestions.

    You’re just deluding yourself if you think you can be using new software that involved zero AI, and that delusion becomes larger and larger for every person involved in it.

    They might not be actively using it, but they have used it, and it has touched the software you use.

    Edit: Also while were at it, ditch your browser you use the internet in, and your phone.

    Edit2: You can continue below if you'd like but I'll short circuit things here.

    Here there’s talk on their repository about using an AI detection service (which is AI) to try and flag AI.

    https://codeberg.org/rimu/pyfedi/issues/605

    https://codeberg.org/rimu/pyfedi/issues/530 and image hashing with a link to his chatgpt chat.

    So ya, AI was used in the development process (planning / discussing / testing) is part of the process.

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 10 child comments
    •
  • [+] -7 points 9 months ago (9 children)

    They’ve used AI too.

    Where?

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 9 child comments
    •
  • [–] 2 points 9 months ago (8 children)

    I literally just gave you all sorts of examples of how or where.

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 8 child comments
    •
  • [–] -3 points 9 months ago (7 children)

    You literally gave me nothing but speculation.

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 7 child comments
    •
  • [–] 3 points 9 months ago (6 children)

    Hence the you're just delusional because you really don't understand what's going on.

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 6 child comments
    •
  • [–] -3 points 9 months ago (5 children)

    You're the delusional one if you think everyone and everything was coded with AI. That's just incredibly ignorant.

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 5 child comments
    •
  • [–] 4 points 9 months ago (4 children)

    I'm not saying it was entirely coded with AI, but AI has had some involvement.

    You clearly stated you think any use should be disclosed, as per the old steam policy, and that you wouldn't use anything if it was disclosed as used.

    In some shape or form, it has been involved with PieFed, and it would require a disclosure if they were being honest about it.

    There are over 40 contributors on codeberg repository. That's 40 people that you have to belive have never asked ChatGPT a question about the work they were doing on it, that didn't base a descision off the Google search result AI response at the top, that didn't use any of the IDE code completion tools built right into the software they use to write the software.

    It's just not happening.

    If you think you can use software like this that isn't tainted by it, you're just flat out wrong at this point and moving forward in the world.

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 4 child comments
    •
  • [–] -2 points 9 months ago (3 children)

    AI has had some involvement

    it has been involved with PieFed

    You keep suggesting this but you can't provide any evidence whatsoever. And until you do I'll continue to reject your nonsense.

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 3 child comments
    •
  • [–] 8 points 9 months ago* (last edited 9 months ago) (2 children)

    Here there's talk on their repository about using an AI detection service (which is AI) to try and flag AI.

    https://codeberg.org/rimu/pyfedi/issues/605

    I tried a 3B version of the same model and it couldn't detect that the sample text that came with the repo was AI. So we'd need a GPU capable of loading an 7B model which means at least 8 GB of RAM. afaik the demo script does two detections as examples of what's possible but either one will do the job.

    They were talking about a service they thought maybe they could even have run on your machine.

    Doesn't matter if they used it or not in the end, AI was used in the development process.

    Tell yourself whatever you need to so you can sleep at night, but of the 40+ people contributing to the project and they're openly talking about AI on their repository, well I can't help you...

    I'm expecting no reply to this message as it would constitute you using a service that involved AI.

    Edit: and pay attention to the names in that thread and who owns the repository.

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 2 child comments
    •
  • [–] 0 points 9 months ago (1 child)

    Doesn't matter if they used it or not in the end

    Uh...that's the only thing that matters.

    AI was used in the development process.

    No it wasn't.

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 1 child comment
    •
  • [–] 1 point 9 months ago (17 children)

    Yes, and while we're at it let'd refuse to read books not written by scribes. And refuse clothes not woven by hand.

    I understand the frustration with the industry, but at the end of the day it is a tool, and it has its uses. Just because it is being misused doesn't mean it's universally bad. This just seems petty and misdirected.

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 17 child comments
    •
  • [–] -3 points 9 months ago (16 children)

    GenAI does not have any uses. It is universally bad. If you think so, you need to stop drinking the corpo KoolAid and pay closer attention.

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 16 child comments
    •
  • [–] 1 point 9 months ago (13 children)

    Then why have seasoned programmers accepted that getting an LLM to generate messy code, then tidying it up, is often faster than writing 2 dozen lines themselves?

    Or myself, I use it with TTRPGs, to simplify NPC creation under a set of structured rules. I still play the characters as unique individuals, but being able to click a button and have 4 personality points to base any nameless NPC around is a lot more fun and dynamic than trying to come up with new characters I didn't expect the group to speak to, on the spot.

    Claiming it does not have any uses at all seems like an expression of your own lack of creativity, or willingness to adapt to new technology. I don't need to worship the tech-bros to find a use for new technologies.

    Can GenAI replace a human? No. There is no context in which human work can be fully replaced by GenAI. But that doesn't mean it cannot simplify and enhance skilled workers that understand its limitations and use it to increase their own productivity.

    Is it possible you're so engrossed in anger and disgust at how it is being marketed, that you're deciding to hate the entire concept rather than the fact it is bring misused, and it's capabilities are being wildly exaggerated to the point of lies? Or that the disgusting manipulation of empty promises on empty promises on empty promises, with the bullshit happening around RAM prices, GPU prices, etc. Or entire workforces being fired in order to be replaced by LLMs is making you prejudiced against the tool, rather than those using it to justify abuse and idiocy?

    If hammers were used to kill more people than guns or knives, would you claim there is no reason to ever use or own a hammer?

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 13 child comments
    •
  • [–] 1 point 9 months ago (12 children)

    Then why have seasoned programmers accepted that getting an LLM to generate messy code, then tidying it up, is often faster than writing 2 dozen lines themselves?

    I don't know that they have.

    The problem is not that it's not faster but that this causes a deep influence on the code, as well as being full of bugs that are glossed over in the thousands of lines of code.

    Claiming it does not have any uses at all seems like an expression of your own lack of creativity

    LOOOOOLOLOL pot meet kettle.

    But that doesn't mean it cannot simplify and enhance skilled workers that understand its limitations and use it to increase their own productivity.

    It can't. It can only be used to generate lazy bug-ridden garbage.

    Is it possible you're so engrossed in anger and disgust at how it is being marketed, that you're deciding to hate the entire concept

    Is it possible that how it's being marketed makes makes you overlook the bug-ridden garbage that it pumps out?

    No, it's possible that I used it and found it to be wrong a majority of the time and that actually just doing it myself is faster and less error-prone. If you only want to pump out trash then by all means. But I don't want any part of it.

    If hammers were used

    I'm so tired of this nonsense. A hammer does not do anything that I do not explicitly make it do. I do not ask a hammer to install the roof and then it runs around driving nails into all the windows.

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 12 child comments
    •
  • [–] 2 points 9 months ago* (last edited 9 months ago) (11 children)

    Why didn't you remark on my own usage? You're speaking from your own experience but seem to be ignoring others. Your personal experience is not more valid than others. Or are you convinced that 100% of people who have found any degree of use for it have some how been tricked?

    Edit.

    Upon re-reading your reply, I have to ask a simple question. What would need to be demonstrated in order to change your mind? If you can think of an honest criteria, we can keep talking about it. If your first response is to say "There is literally nothing that can change my mind." Then this is not a discussion, it is simply you expressing anger and indignation.

    And I'm sorry to say, but if you're engaging in conversation without a single iota of willingness to see the other sides perspective or reasoning, then you're a bad conversation partner and are consciously choosing to be arrogant, even if your side is the "Correct" side.

    Have I come across as rude or dismissive that you felt the need to mock and belittle me? I tried to form my response respectfully while pointing out possible areas of bias.

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 11 child comments
    •
  • [–] 0 points 9 months ago* (10 children)

    Why didn’t you remark on my own usage?

    Why would I comment on your usage?

    You’re speaking from your own experience but seem to be ignoring others

    I am not ignoring anything, I just literally cannot speak to your experience. I can only speak to mine.

    Or are you convinced that 100% of people who have found any degree of use for it have some how been tricked?

    They've been grifted, yes. I know they have because I've had these conversations over and over again. They come to me with some sort of untrue statement. I ask them where they got it. They say Google. I ask them if it's the AI Overview and they say yes. I ask them where the AI got that information. They have no idea. I use a real search engine, find an authoritative source that directly and clearly refutes their statement, and they're confused as to how I got it. They are completely ignorant to the fact that it's constantly wrong and that it's use is literally making them dumber, not only because the answer is wrong, but because they're not actually exercising their brain to find it.

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 10 child comments
    •
  • [–] 2 points 9 months ago (9 children)

    You should speak to others experiences because you are making the universal statement that is has literally no uses, contrary to many people stating they have, in fact, found uses for it.

    So either every single person that believes they have found a novel use for it is wrong. Or you have universally decided that none of their experiences are valid in forming your opinion.

    Considering I have found a use for it, that does not require it to write code, paint pictures, or tell me I'm right about everything, why is my usage invalid in nullifying your statement that "GenAI does not have any uses."

    There is no ambiguity in that statement, and yet I have found use for it.

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 9 child comments
    •
  • [–] 0 points 9 months ago (8 children)

    You should speak to others experiences

    My brother in Christ, what did I just say? I literally cannot. I don't know anything about your experiences. They're your experiences. And I just finished explaining exactly how people's experiences are invalid because they don't understand what's happening. What do you want?

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 8 child comments
    •
  • [–] 2 points 9 months ago (7 children)

    You're contradicting yourself and not seeing it. You're universally saying it has absolutely no uses, and using your own experiences and others, as evidence. When counter-evidence is provided, you dismiss it because you can only speak from your own experiences.

    Either other peoples experiences are valid, and you must accept that some people have found genuine uses for the technology, despite your hatred for the industry and the false marketing around it.

    Or other peoples experiences are not valid, in which case there's no point in talking about anything, because you will not consider it valid unless you personally experience it.

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 7 child comments
    •
  • [–] 1 point 9 months ago (5 children)

    When counter-evidence is provided, you dismiss it because you can only speak from your own experiences.

    Your experience is not evidence. It's an anecdote.

    in which case there’s no point in talking about anything, because you will not consider it valid unless you personally experience it.

    In a world where everyone and their fucking mother has been trying relentlessly to gaslight me into believing this for 3 fucking years, despite any and all available evidence, yes, I absolutely will not. I'm long past done entertaining the notion.

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 5 child comments
    •
  • [–] 1 point 9 months ago (4 children)

    Actually, that's materially incorrect. You are making the universal claim that a technology has literally no uses, and every single person that claims otherwise is either ignorant, or has been grifted.

    I am making the counter-argument that, while I agree with the overall concerns with misuse, and misrepresentation, claiming it has literally no use is objectively false.

    Anecdotal would be "I've heard people have found uses for it." Or "My cousin says it helps him do XYZ."

    I have given you a specific example of my own use. That is not anecdotal, that is a definitive, replicateable, counter point.

    You have decided to dismiss it, because making a universal sweeping claim inherently requires the dismissal of all counter points in order to remain intact.

    I'm sorry that you're so angry, and have convinced yourself the world is trying to gaslight you, that doesn't seem like a healthy belief. I won't be responding further.

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 4 child comments
    •
  • [–] 1 point 9 months ago* (2 children)

    Just FYI, at this point this guy has been provided 2 links on the Pyfed codeberg repository that show AI was used in the development of PieFed and he says he won't use anything that did, so any further replies from him make him a hypocrite for using software that used AI while being developed. He's a troll/wasting your time.

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 2 child comments
    •
  • [–] 2 points 9 months ago (1 child)

    Yeah, I figured that was probably the case when I asked what evidence might change their mind, and the question was just ignored. Still, I was mildly entertained by pointing out their logical errors, so it wasn't a complete waste of my time lol.

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 1 child comment
    •
  • [–] 0 points 9 months ago

    I am making the counter-argument that

    No you're not. This claim has nothing to do with your personal experience. You are citing your personal experience as evidence. You're doing the exact same thing you're accusing me of, only I have provided actual evidence to back up my statements, and you haven't.

    I have given you a specific example of my own use. That is not anecdotal

    ...yes? That's the definition of anecdotal.

  • permalink
  • fedilink
  • source
  • parent
  • [–] 1 point 9 months ago

    Just uninstall all games made after 2022 then, because I can assure you llm' have been used for code in some capacity in every game. But I would argue there is a big difference in using AI for assert generation. And using it to help read docs or getting ideas for refactoring some code etc

  • permalink
  • fedilink
  • source
  • parent
  • [–] 0 points 9 months ago

    Are you opposed to, for example, AI being used to bug fix?

    Personally my opposition is mostly in the form of drawing art, writing plotlines, recording voicelines, etc. I am not opposed to AI being used in certain aspects of game development.

  • permalink
  • fedilink
  • source
  • parent