cross-posted from: https://beehaw.org/post/28095114

Vapor World: Over The Mind is a Soulslike Metroidvania with an evocative, painterly art style reminiscent of Lies of P and the narrative platformers of Playdead. According to Korean studio Alive Inc, the game had 100,000 wishlists just before its early access launch on August 19, mostly on the strength of a demo available since January 2025.

But what should have been a triumphant early access launch went pear shaped very quickly. At launch, few talked about the quite distinctive worldbuilding or the parry-heavy combat. Instead, all focus was trained entirely on something absent from the popular demo: very obvious AI use.

you are viewing a single comment's thread
view the rest of the comments
[–] 83 points 1 month ago* (76 children)

"The truth is, we spent a long time trying to tell this game's story with cutscenes made entirely by our own hands," the statement goes on. "The results were devastating every time. We came to believe we could not properly deliver this game's complex, difficult narrative that way, and that is why we turned to AI. But if that choice hurt your experience, the fault is not in the tool. It is in our own shortcomings."

Man, if this is true – that they agonized over the cutscenes but decided they couldn't and that AI would be better* – then I feel legitimately bad for them.

I went in assuming it was strictly a lazy "ah, fuck it" kind of thing, but even though this still wouldn't be acceptable, it seems like this team really is hard-working, cared about the final product, and actually did think AI could deliver a good experience. Deluded, but that seems like part of a larger culture of delusion in SK.*

* Notably, the article remarks on how sentiment around AI in South Korea is vastly more positive than e.g. the US.

  • source
  • hideshow 76 child comments
  • [–] 73 points 1 month ago (6 children)

    That's basically a statement for the press to try to minimize damage. He even said they had the cut scenes done and later replaced them with AI. Bullshit.

    Unless AI takes one more leap and reaches a point where no one can tell the difference, what I believed was still closer to an ideal than a reality.

    They are mad they got caught and it's the AIs fault for not being good enough to hide it.

  • source
  • parent
  • hideshow 6 child comments
  • [–] 11 points 1 month ago* (last edited 1 month ago) (5 children)

    it's the AIs fault for not being good enough to hide it.

    "if that choice hurt your experience, the fault is not in the tool. It is in our own shortcomings."

    The director literally outright says it's their fault for using the tool.

    And what do you mean "got caught"? It's very obvious. They mistakenly thought most audiences would enjoy it, not that audiences wouldn't notice. Which makes sense given the overwhelming difference in the perception of AI between SK and the West. He even outright says he expected people to notice: "I expected some of you would be disappointed by our use of AI. I did not expect it to become this big."

  • source
  • parent
  • hideshow 5 child comments
  • load more comments (4 replies)
  • [–] 43 points 1 month ago (11 children)

    I'd rather see the "bad" things than Ai.

    Honestly if the rest of the game is good, I would rather have some napkin sketches, or a computer based paint software.

  • source
  • parent
  • hideshow 11 child comments
  • [+] 5 points 1 month ago* (last edited 1 month ago) (6 children)
  • [–] 4 points 1 month ago (3 children)

    "Sentinels of the Multiverse" is a fairly popular board game (within those circles) and it looks pretty rough. Pokemon S/V got rightfully dunked on for looking like shit but it still made a bunch of money. "The Case of the Golden Idol" is highly regarded. Runescape.

  • source
  • parent
  • hideshow 3 child comments
  • [–] 2 points 1 month ago

    No one really gave pokémon any s*** for the art. They gave s*** for the performance such as like the pop-in and reused assets.

    But the actual art is fine. No one really had issue with that. It's pokémon for heaven's sakes. Not crisis.

    That's why people got mad. It's basic 3D models in really simple backgrounds and it ran like s***. There's no reason for something that looks like pokémon to run poorly.

  • source
  • parent
  • [–] 2 points 1 month ago (1 child)

    I think pokemon was so heavily criticized partly because it's developed by a veteran studio that really should be able to do better. And the way it looked bad wasn't so much the art style, but how it was loaded into the game.

    You had a lot of pop-ins, things not being animated in the distance, and every npc using the exact same animation at the exact same time, as well as some truly uninspired textures out in the wild.

    Gamefreak is clearly capable of better when you look at Beast of Reincarnation or Little Town Hero, they just don't seem to care about pokemon anymore.

  • source
  • parent
  • hideshow 1 child comment
  • load more comments (1 reply)
  • load more comments (3 replies)
  • [–] 17 points 1 month ago* (3 children)

    Why not hire people or another firm to do the cutscenes then. You know, professionals who have learned and perfected their craft. Professionals you don’t have.

    Backslash shouldn’t have been a surprise.

  • source
  • parent
  • hideshow 3 child comments
  • [–] 10 points 1 month ago* (3 children)

    Still seems like a cop-out.

    If they had intended to do it manually in the first place, I don’t see why they couldn’t just use AI to help generate a number of rough outlines to help conceptualize a solid template that satisfied their needs before investing their time into it.

  • source
  • parent
  • hideshow 3 child comments
  • [–] 3 points 1 month ago* (last edited 1 month ago)

    AI could maybe help stress test the game before the final release. If AI were used like that, the game itself would likely not be considered slopified.

    The dirty data centers, the RAM chip grab, the widening divide between the haves and the have nots being exacerbated by the billionaire AI ownership, these problems have to be addressed before I can even consider the other issues. AI as tech is usable, but as a socioeconomic phenomenon it is hugely problematic.

  • source
  • parent
  • load more comments (2 replies)
  • [–] 6 points 1 month ago

    Man, if this is true – that they agonized over the cutscenes but decided they couldn't and that AI would be better* – then I feel legitimately bad for them.

    even then, you use the AI output as inspiration, not as the end product..

  • source
  • parent
  • [–] 4 points 1 month ago (2 children)

    Interesting context, but that statement sounds so fucking stupid.

    "Welp, we tried so hard, but just couldn't portray our art in the best possible way. So naturally, we turned to the soulless machine that fails at generating anything resembling art even if you tell it to do a 1:1 copy of a real art piece."

  • source
  • parent
  • hideshow 2 child comments
  • [–] 3 points 1 month ago (2 children)

    I was thinking the same but I have so many questions (that I don't care enough to research).

    How big is their team? Could they not find any competent artists? Could they not afford any competent artists? Could they not afford to upskill any of their current team, like getting them lessons in whatever was relevant?

    It seems like they were just shallow and wanted a "classically" pretty product right now. And that doesn't engender much sympathy.

  • source
  • parent
  • hideshow 2 child comments
  • [–] 1 point 1 month ago (30 children)

    It's almost as if neither "ai can do everything for me" nor the ai witch hunting are healthy approaches, but, as always, people keep polarizing themselves to hate on the opposite tribe.

    I'm so fucking tired of this bullshit...

  • source
  • parent
  • hideshow 30 child comments
  • [–] 4 points 1 month ago (1 child)

    The pendulum has only just started to swing. Homeostasis will take time.

  • source
  • parent
  • hideshow 1 child comment
  • [–] 9 points 1 month ago*

    Unfortunately it may be politicized beyond recovery, at least in the US with the trajectory of things.

    Its going to be associated with the corporations doing awful things with it, and thats going to blot out the sun.

  • source
  • parent
  • [–] 3 points 1 month ago (14 children)

    Yyyyyyyyup. Even attempts at honest discussion get polarized. "If you use AI(LLMs) you're an unethical AI-Bro!"

    Oh... I didn't realize that my self-hosted open source small model trained on Wikipedia and Stack Exchange was unethical. Thank you for showing me the error of my ways.

  • source
  • parent
  • hideshow 14 child comments
  • [–] 6 points 1 month ago* (12 children)

    Yes, you are acting unethical. Your self-hosted model is neither open nor was it trained with consent. Using them is just as unethical as using the models from OpenAI.

  • source
  • parent
  • hideshow 12 child comments
  • [–] 7 points 1 month ago (7 children)

    CC BY-SA gives freedom of use. Why do you think additional consent would be necessary? Or do you mean that because the restricting conditions are not met, additional consent would be necessary?

    Given the form of transformation, Share-Alike is certainly difficult to meet, if not impossible. Is that what you referred to with "is not open"? Or do you mean something else?

  • source
  • parent
  • hideshow 7 child comments
  • load more comments (7 replies)
  • [–] 6 points 1 month ago*

    Some models do use open datasets and largely reproducible training, with the Nvidia Nemotron series being the highest profile example. You are right that many are opaque about training data, though, and this is not okay.

    Using them is just as unethical as using the models from OpenAI.

    And I vehemently disagree with this.

    OpenAI is a different order of magnitude of moral depravity. Thats like saying using Lemmy is just as bad as using Facebook or Palantir.

  • source
  • parent
  • load more comments (3 replies)
  • load more comments (1 reply)
  • load more comments (10 replies)
  • load more comments (8 replies)