AI Driving Game Players Are Burning Through So Many Tokens Its Developer Had to Take Out a Bank Loan to Keep It Running

Comments

• Rhynoplaz@lemmy.world ⁨2⁩ ⁨weeks⁩ ago

  How can this ever be sustainable without a huge monthly subscription?

  source

  • Katana314@lemmy.world ⁨2⁩ ⁨weeks⁩ ago

    Ask Anthropic! I hear they worked out a solution.

    source

    • VibeSurgeon@piefed.social ⁨2⁩ ⁨weeks⁩ ago

      The solution being a huge monthly subscription. Or massive token rates if you’re a big customer

      source
  • Babalugats@feddit.uk ⁨2⁩ ⁨weeks⁩ ago

    They keep borrowing and giving out ‘free’ tokens until people are hooked.

    Then gradually intrusive different payment methods until they figure out what works best.

    Then prices rise (I’m guessing, rather rapidly) to pay off all of the loans and to start making profits.

    In the meantime, their machines are learning from the users, and learning which areas and how it can be most profitable.

    Boom. They’ve created a profitable business that nobody wanted, nobody asked for and nobody needs. The interactors don’t even realise that they’re not even really using it, there training it.

    DON’T USE AI.

    YOU’RE CREATING MORE PROBLEMS. MORE BEZOS’ AND MUSKS.

    source
  • Creat@discuss.tchncs.de ⁨2⁩ ⁨weeks⁩ ago

    It can’t. But somehow the dev hasn’t noticed I guess?

    source

    • Rhynoplaz@lemmy.world ⁨2⁩ ⁨weeks⁩ ago

      When you’re taking out loans you’ve noticed.

      source

      • -> View More Comments
  • sp3ctr4l@lemmy.dbzer0.com ⁨2⁩ ⁨weeks⁩ ago

    Easy, you just hyperinflate the USD!

    Infinite money glitch, irl.

    source
• FridaVanMercury@discuss.tchncs.de ⁨2⁩ ⁨weeks⁩ ago

  From their blog on Steam:

  For the paid release, we plan to account for expected AI usage in the game’s price. Players will not be charged separately for their individual AI token usage after purchasing the game. This isn’t because each player is particularly expensive to support. The total cost has increased because so many people are playing, and we have taken out a bank loan to help keep the demo running.

  I do not understand. They have to factor in token usage to the game price, but also the longer people play the more they will lose money over time? How is that even an operating model.

  source

  • Grimy@lemmy.world ⁨2⁩ ⁨weeks⁩ ago

    If they don’t let players hook up their own API key, it’s a guarantee they shit it down after a year.

    source

    • PHLAK@lemmy.world ⁨2⁩ ⁨weeks⁩ ago

      More like a month.

      source
    • Cornballer@lemmy.zip ⁨2⁩ ⁨weeks⁩ ago

      It’s an avenue for abuse. Buy the game: unlimited “free” tokens. Like happened with Chipotle’s AI.

      source
  • a_non_monotonic_function@lemmy.world ⁨2⁩ ⁨weeks⁩ ago

    My guess is that they are just morons. It is really easy to be confident in your projections if you are a stupid person.

    source
  • Wispy2891@lemmy.world ⁨2⁩ ⁨weeks⁩ ago

    It seems that it’s one of those games with zero replayability

    They probably bank on streamers make fun videos in 4-5 hours of gameplay total, then the people that buy the stuff and never play it

    Btw it uses Gemini and gpt, it’s expensive to run that.

    What if someone pirates the game and starts wasting tokens?

    source

    • Bane_Killgrind@lemmy.dbzer0.com ⁨2⁩ ⁨weeks⁩ ago

      zero replayability

      Missing one of the only good reasons to include AI in your game…

      source
    • SkunkWorkz@lemmy.world ⁨2⁩ ⁨weeks⁩ ago

      And a 4chan troll army could bankrupt them for funsies.

      source
  • rethnor@lemmy.zip ⁨2⁩ ⁨weeks⁩ ago

    Ai probably approved the idea and said it was great.

    source
  • GenosseFlosse@feddit.org ⁨2⁩ ⁨weeks⁩ ago

    People won’t play forever. Even good games are only played for maybe 100h before some other new game comes along. If you know the average game time you can calculate the ai costs.

    source

    • BlaestEgnen@feddit.dk ⁨2⁩ ⁨weeks⁩ ago

      You’re still opening up for credentials to be harvested and troll requested by script kiddos

      source

      • -> View More Comments
  • ILikeBoobies@lemmy.ca ⁨2⁩ ⁨weeks⁩ ago

    Loans are normal and you take losses to capture the market then you enshittify

    source

    • aksdb@lemmy.world ⁨2⁩ ⁨weeks⁩ ago

      But loans don’t scale with the amount of players and played hours. They purely factor into the cost of development and support.

      source

      • -> View More Comments
• Funkt4st1c@lemmy.world ⁨2⁩ ⁨weeks⁩ ago

  You could literally spend $0 and an afternoon hooking up a 4gb model to the game and having it run locally instead. Or you could also spend $0 and code a decision tree that would be fun for players to experience. Or you could choose the dumbest possible option and bankrupt yourself for no reason

  source

  • p03locke@lemmy.dbzer0.com ⁨2⁩ ⁨weeks⁩ ago

    Why are there so many opportunities to post this meme?

    We are not the same!

    source

    • DummyWOW@lemmy.zip ⁨2⁩ ⁨weeks⁩ ago

      lol not reading that

      source

      • -> View More Comments
    • Funkt4st1c@lemmy.world ⁨2⁩ ⁨weeks⁩ ago

      No you dont.

      source
    • ddplf@szmer.info ⁨2⁩ ⁨weeks⁩ ago

      Boooy is this “meme” exposing your well deserved insecurities.

      source
    • keithnator3000@lemmy.world ⁨2⁩ ⁨weeks⁩ ago

      There will only be more as time passes.

      source
    • magic_smoke@lemmy.blahaj.zone ⁨2⁩ ⁨weeks⁩ ago

      The only model I program with is a Model M.

      Well maybe its a Mini M, but close enough XP

      source
  • Aneb@lemmy.world ⁨2⁩ ⁨weeks⁩ ago

    Number 3 please!

    source
• iceberg314@slrpnk.net ⁨2⁩ ⁨weeks⁩ ago

  Why can’t they just fine tune a small locally hosted model? Are they dumb?

  Most gamers have a GPU anyways

  source

  • Bluefruit@lemmy.world ⁨2⁩ ⁨weeks⁩ ago

    Exactly my thought. Surely you could and achieve at least a similar result.

    I would say yes, yes they are dumb. Taking out a loan for a game that isn’t even out yet according to the article to keep an ai going is pretty stupid. Even if the game does well, the overhead is massive for what could be done with a smaller specialized model. Hell, I bet payig someone to train an existing model would be significantly less than a $1000 a day when you aren’t even making money yet.

    source
  • DrPop@lemmy.world ⁨2⁩ ⁨weeks⁩ ago

    Because this is vibe-coded. You think an LLM is going to give up how to make your own?

    source

    • FunnyUsername@lemmy.world ⁨2⁩ ⁨weeks⁩ ago

      yes. it’s not self aware.

      source
    • CybranM@feddit.nu ⁨2⁩ ⁨weeks⁩ ago

      I dislike AI slop as much as the next guy but that’s a stupid take

      source
    • boonhet@sopuli.xyz ⁨2⁩ ⁨weeks⁩ ago

      I literally let deepseek + opencode modify the nanogpt code for my own custom model and suggest training datasets lol

      Ran out of patience after a while because crappy GPU, but it was training and words were being worded when I prompted it after the first day of training

      source
    • chicken@lemmy.dbzer0.com ⁨2⁩ ⁨weeks⁩ ago

      It will though

      source
  • GenosseFlosse@feddit.org ⁨2⁩ ⁨weeks⁩ ago

    Not everyone has a GPU with enough RAM. I have a 3060 with 12GB, and I cannot run ollama and a 3D game at the same time. As soon as ollama starts processing, the GPU is at 100% and game and ollama start swapping data in and out of the GPU RAM, which means the whole computer becomes unresponsive for a long time. I had to set up a script that disables ollama if steam is running.

    source

    • iceberg314@slrpnk.net ⁨2⁩ ⁨weeks⁩ ago

      12GB is a lot of ram. By small local models I am talking about fine tuning something like < 2GB LLM

      source
    • esc@piefed.social ⁨2⁩ ⁨weeks⁩ ago

      1. ollama is bad at everything they do. 2. They can run a really small llm, as the requierements are pretty limited.

      source
    • OrteilGenou@lemmy.world ⁨2⁩ ⁨weeks⁩ ago

      Sorry, all of your what?

      source
• Wakko@hell.cloud ⁨2⁩ ⁨weeks⁩ ago

  “Markets can remain irrational longer than you can remain solvent.”

  source

  • CrimeIsLegalNow@ani.social ⁨2⁩ ⁨weeks⁩ ago

    One of the important things to understand is that certain elements of that irrational market are very, very rational, as long as the money comes in.

    It’s what happens when the chumps you’re grifting either smarten up or go broke that things get bumpy.

    source
• SkunkWorkz@lemmy.world ⁨2⁩ ⁨weeks⁩ ago

  So they didn’t even do the math for a basic cost forecast before launching the demo. Also this shit makes them vulnerable for bad actors. Just put this game on several PCs and run it 24/7 to bankrupt them.

  source

  • ColeSloth@discuss.tchncs.de ⁨2⁩ ⁨weeks⁩ ago

    What makes you think that the demo just got massively more popular than expected? This is probably an occam’s razor type of thing.

    I’m more curious about how they were planning on structuring the full game to make a profit. Hopefully they’re just planning on using a client side ai, but that would likely make the demo somewhat deceitful.

    source

    • Valmond@lemmy.dbzer0.com ⁨2⁩ ⁨weeks⁩ ago

      Minimum Specs : 96GB VRAM

      source

      • -> View More Comments
• toxicbubble@lemmy.world ⁨2⁩ ⁨weeks⁩ ago

  me explaining to future generations why all the water dried up

  source

  • goferking0@lemmy.sdf.org ⁨2⁩ ⁨weeks⁩ ago

    And why computers became rentals again

    source
• Mordikan@kbin.earth ⁨2⁩ ⁨weeks⁩ ago

  Another engineer I work with had a really well put description of proper AI implementation that really applies to this:

  "AI should help you build the automation, AI should not be the automation."

  source

  • Valmond@lemmy.dbzer0.com ⁨2⁩ ⁨weeks⁩ ago

    AI should help, not do.

    source

    • SkunkWorkz@lemmy.world ⁨2⁩ ⁨weeks⁩ ago

      AI: only do, not help

      source
• grahamja@reddthat.com ⁨2⁩ ⁨weeks⁩ ago

  AI running a video game, somehow still less power efficient than a playstation 1. Good job guys. Just make an actual video game.

  source
• MangoPenguin@lemmy.blahaj.zone ⁨2⁩ ⁨weeks⁩ ago

  Given that people playing this have GPUs, why not run the AI model locally? I can’t imagine it needs to be a very large model.

  source

  • aksdb@lemmy.world ⁨2⁩ ⁨weeks⁩ ago

    Actually sounds like the perfect use case for a model like Jev. I think Jevos should be able to handle that locally.

    source
  • EncryptKeeper@lemmy.world ⁨2⁩ ⁨weeks⁩ ago

    Because even very small models will take up a large chunk of your cards VRAM, if not all of it. And then you still have to render the game on top of it.

    source

    • MangoPenguin@lemmy.blahaj.zone ⁨2⁩ ⁨weeks⁩ ago

      It looks like a very simple game graphically, so having say 2GB VRAM left over for a small model would probably be viable?

      source

      • -> View More Comments
    • humanspiral@lemmy.ca ⁨2⁩ ⁨weeks⁩ ago

      There are 1-2gb vram models that should fit in anything. Game doesn’t appear to be texture heavy, and targetting a 6gb gpu, would cover anything less than 10-12 years old.

      source

      • -> View More Comments
• lepinkainen@lemmy.world ⁨2⁩ ⁨weeks⁩ ago

  AI powered games don’t work.

  We did a super cool project for a game jam last year. Cost like 50c per game for us.

  Cool for the 100 or so players who tried it, but there’s no way to make a profit from it.

  Local models maybe.

  source

  • Rhaedas@fedia.io ⁨2⁩ ⁨weeks⁩ ago

    For just simple and guided actions and dialogue, local would be fine for this. Or better, have AI design the large database of what to say or do when a player does things. No need to reinvent what players are going to duplicate. The token cost was worse when you consider all the same stuff being prompted over and over.

    source

    • Ledivin@lemmy.world ⁨2⁩ ⁨weeks⁩ ago

      If they’re actually hitting a cloud LLM, there’s no good reason not to wrap that and memoize the results… it would probably be hard to generalize the inputs (it could just be the whole game state for all we know 🤷‍♂️) but youd build up that database of all the most common input/outputs in real-time

      source

      • -> View More Comments
  • HeHoXa@lemmy.zip ⁨2⁩ ⁨weeks⁩ ago

    In case it might be interesting info, maybe want to check out WebGPU, WebLLM, Wllama, or Transformers.js

    You can write web apps that run an llm on the client’s browser. It requires permission and compatibility and isn’t… it might be tough to really make accessible to the average user.

    But it’s an option and kinda cool.

    source
• Cricket@lemmy.zip ⁨2⁩ ⁨weeks⁩ ago

  How long before the internet is flooded with clips of the little sister saying all kinds of nasty stuff after people prompt inject the AI?

  source
• Davel23@fedia.io ⁨2⁩ ⁨weeks⁩ ago

  The AI disclosure on Steam claims the sister's voice is generated by a local model. If they're being charged for tokens I somehow doubt that.

  source

  • SkunkWorkz@lemmy.world ⁨2⁩ ⁨weeks⁩ ago

    Probably only the voice is local. Natural voice requires a voice synthesis model. The “reasoning” is probably offloaded to a cloud service.

    source
• frustrated_phagocytosis@fedia.io ⁨2⁩ ⁨weeks⁩ ago

  That sounds like the most boring game imaginable and I hope they go bankrupt from it

  source

  • Kolanaki@pawb.social ⁨2⁩ ⁨weeks⁩ ago

    Huess you’ve never heard of Desert Bus.

    source

    • frustrated_phagocytosis@fedia.io ⁨2⁩ ⁨weeks⁩ ago

      Does Desert Bus require tokens? If not, carry on.

      source
    • Rhaedas@fedia.io ⁨2⁩ ⁨weeks⁩ ago

      I have. Sometimes boring has a weird attraction.

      source

      • -> View More Comments
    • Widdershins@lemmy.world ⁨2⁩ ⁨weeks⁩ ago

      Desert Bus was part of a compilation of six games and if you played it today it would still work.

      source
• bitjunkie@lemmy.world ⁨2⁩ ⁨weeks⁩ ago

  Steam link

  It’s every bit as horrible as you’re probably imagining. She has one of those talks-too-fast fake slop voices like on the YouTube spam shorts.

  source
• chunes@lemmy.world ⁨2⁩ ⁨weeks⁩ ago

  Hilarious article to put out tbh. If the game wasn’t broken before, it’s gonna be now.

  source
• trashboat@piefed.social ⁨2⁩ ⁨weeks⁩ ago

  And they say there’s a bubble, pffffffffffffffffft

  source
• Pratai@piefed.ca ⁨2⁩ ⁨weeks⁩ ago

  Good. Go bankrupt asshole.

  source