32
Big Walk Fixed a Bug With AI (thelemmy.club)
submitted 6 hours ago by to c/games@hexbear.net
30 comments fedilink hide all child comments
 

Some awards show controversy had House House disclose possibly the lightest possible LLM usage you can imagine. Anyway it's a Bad Game now.

Amongst all the glut of genuinely vibe coded slop, inventive AI reverse engineering and the plain ol' realities of LLM living in modern software development, I get a bit frustrated seeing it all grouped into one essentialized category.

The common understanding of what AI even is, is at an all-time low, imo. People see 'AI code' next to a game title and imagine someone prompting claude for 20 minutes, and these sorts of stories and over-the-top reactions don't help.

The basics of what happened with this story, is that a particularly bad Unity bug plagued the team and the unity support wasn't being a help. LLM fixed the bug. I'm shocked to see that even this usage is considered unacceptable to some people.

you are viewing a single comment's thread
view the rest of the comments
[–] [S] 4 points 4 hours ago* (5 children)

I don't disagree with the critique of the technology, though I think the realities of it are a lot less poetic. Under current property dynamics it's almost purely an extractive technology that has more or less stolen collective productivity and turned it into a slop machine, but the thing still exists and is capable of production. This is less of a 'cats out of the bag' appeal and more 'we're going to have to seize control of this technology just like everything else' line.

To that end, when I imagine an ideal possible scenario where this technology now has to exist alongside fundamentally different market dynamics, I have to imagine the application of local models in specialized fields would be one of the ways it sticks around.

And, I think, it will stick around - we have to meet people where they are at a certain point, and software devs range from vibe code addicts to professionals indulging in a powerful new tool. None of them want to give it up, I don't think.

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 5 child comments
    •
  • [–] 10 points 3 hours ago* (last edited 3 hours ago)

    Some technologies are stupid and bad actually. The stuff people often say they're useful for that are on the scale of tasks that reasonable models could be built for are tasks that more efficient software already exists to solve. Like you don't need some nuts gpu full of conflict minerals and 500 joules of electricity to rename some variables or adjust the tone of an email. You should go and read the papers that describe how to build these things, make a small language model, see how stupid of an approach it is. They're dead simple just really big and inefficient.

    Many software devs hate this shit. I am a dev and I hate this shit. It has absolutely destroyed the pipeline for training people in good ethics and practice and is actively destroying what little good practice we have managed to cultivate in software design. This is decent reading: https://tante.cc/2026/04/21/ai-as-a-fascist-artifact/ and so is this: https://deadsimpletech.com/blog/no-such-thing-as-just-a-tool

  • permalink
  • fedilink
  • source
  • parent
  • [–] 7 points 3 hours ago

    also go and read the transcripts of microsoft standing in front of the EU parliament talking about how we need to blow past climate goals for AI. This isn't some stupid bougie concern, millions of people are going to die.

  • permalink
  • fedilink
  • source
  • parent
  • [–] 5 points 4 hours ago (1 child)

    I'm in the "professional with a powerful new tool" category, using copilot in an IDE on a work project for the first time.

    Been messing with code for over 10 years but never in a professional capacity and it's been helpful for quickly getting me up to speed with unfamiliar libraries or languages, or HW interfaces. I'd be fine giving it up tho.

    It's an undeniably powerful tool with genuinely impressive capabilities. Because of this I'm convinced that a zero tolerance policy is a losing proposition long term in terms of popular support.

    But it's also an open question what the real cost of operating these things is in the long term is. We're still in the bubble with huge amounts of capital going around to make these available for use below their actual cost. It's likely many would not fine it worth using if presented with its unsubsidized cost.

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 1 child comment
    •
  • [–] 7 points 3 hours ago

    Because of this I'm convinced that a zero tolerance policy is a losing proposition long term in terms of popular support.

    i think this is also a big one. any reasonably informed historical perspective has to contend with the 50 movements against automation over the past 2 centuries and how they have all near universally failed. theres plenty of blame not due to those within these movements, but at some point you have to contend that "harassing those who use a technology to force it to recede" is not a viable strategy given the conditions at hand. and given the average argument actively reifies intellectual property or goes on about True Art and The Human Soul, i dont think this will amount to much either.

  • permalink
  • fedilink
  • source
  • parent