My take on Big Walk's AI disclosure (https://bigwalk.game/faq/#ai), because I love this game and that the team chose to make it. I disapprove but sympathise with them. My position on LLM use is clear - I hate what they're doing to the industry, and they're built entirely on stolen work - but I have asked myself this question: what if, near the end of development, I had some really horrid core engine problems I couldn't solve that were ruining the game? Could I resist asking a chatbot at last resort?
0 Boosts
0 Likes
@klord Yeah. I do get the sentiment that they should have done better, and they were well funded, they should have found another way etc. But hindsight is great and all vs when you're in the shit spin cycle - I simultaneously disapprove and think they should get a bruised rep over this, but also sympathise at the same time.
@sinbad tricky. If the dates locked in, doesn’t matter how much cash you have. I’ve been in that situation a few times, and it’s fucking horrible.
@sinbad also in a world where a google search may just decide to use an LLM to answer, we have to understand a baseline poisoning is already happening.
Big Walk is a Unity game, the engine itself is full of AI code at this point. Do we count that too?
@cuirass_entertainment @sinbad yes, but using an older version (and choosing not to take it's patches) is not an option really for devs releasing on consoles, as they must update to get the SDK changes needed to publish or patch the game on those platforms.
All releases out of unity in the year or so almost certainly include agentic programming, especially the maintenance patches, which is a much bigger transgression that what Big Walk talk about here.
@mdiluz I think, given the mass adoption in the software industry, you can't really be held responsible for other people's choices, particularly after you've committed to a library (sometimes you can extract yourself from it, sometimes you can stick to earlier versions, but it's not always possible). Only your own
I hope that I could, because even using these tools as a last resort means benefiting from the mass theft and planet burning energy usage, and even on a small level, reinforces their usage and undermines my objections. But if I really really couldn't solve it after weeks of effort? I hope I could resist. But despite my strong objections I wouldn't toss this team into the bone pit for it, especially if they've learned from it. I respect people's choice to not buy it / recommend it now though
I'd like to think, knowing myself, that I'm a stubborn enough old bastard to relentlessly pore over the UE code until I fixed whatever it was, regardless of how long it took or how nasty the workaround ended up being. I've never intentionally fired up a chatbot, ever (barring being ambushed on search engines), much to the delight of the AI-pilled who want to use that as a gotcha. So I think I wouldn't fall off the wagon here. But, respecting humans means understanding our weaknesses too
I think this is a good take with more information than I had on what actually happened: https://blog.lauramichet.com/some-igf-jurors-released-a-statement-about-ai-usage/
@sinbad "we were able to speak directly to an engineer at Unity, who asked around and ultimately concluded that what we wanted wasn’t possible."
FFS, I bet they paid Unity for that response, as well. :(