At this point using any kind of generative “AI” in game development is utterly self-defeating. I’ve lost interest in that whole mess until it stops being a resource-sucking plagiarism machine, which it’s probably going to keep being by definition. I’ve always done coding myself. I don’t even ask the “Claude” thing whatever that is, I just use regular web searches without that whole automatic chatbot summary on top for things like Stack Overflow as I always have been, or maybe the MDN Web Docs if it’s an HTML or JavaScript thing. Then there’s whatever IntelliSense thing Visual Studio and later VSCode or its branches includes that never required a chatbot connection, which can be hit or miss but has mostly worked. The most I might use for automation is what was in use before the whole “vibe coding” enema happened, things like build scripts for programs, or scripts I wrote in Python to quickly format pages on this site that all share the same basic template, like this notes section or the review blog. I could have content like that dynamically load in JavaScript, and on some specific pages with a lot of data I do, but for the most part I stick with static pages for general accessibility.
Honestly, part of this aversion to relying on chatbots is due to dealing with certain work at my job requiring confidentiality, which means keeping materials, including code, from leaving the places it belongs, and not knowing what online chatbots or “integrated agents” are exactly doing with data, just assuming they’re stockpiling it to either sell or accidentally leak or get hacked and leaked that way. Essentially even if my work forced me to use some generative “helper” I’d have to make sure it’s all self-contained and offline. I’ve looked at the options for that and they’re still not great, and I don’t see any point in myself investing time into figuring them out. It’s something I already at least mostly know how to do, so why relearn the whole procedure to have something else do it worse?
So my initial point is that due to the explosion of the tech bro circlejerk that is “AI” in general, technology in general is quickly becoming unaffordable to anyone. My prediction is within a couple years, provided society doesn’t totally collapse and reject any and all technology except guns, a basic “grandma e-mail” PC will instead be a terminal connecting to only a chatbot accepting prompts to do anything with its cloud instance under full screen surveillance, and it will also require a steady supply of human blood for “verification” meaning a direct line with a needle would be placed into the user’s veins, which is also used to run the machine in the first place due to requiring some arcane form of blood magic in place of the usual silicon technology that would go in there but was completely occupied by the data centers right outside the window. The up-front cost will also be the equivalent of $10000 USD in today’s money, not adjusted for future hyperinflation, with a monthly $1000 USD fee on top. And this is even accounting for if or when the whole circlejerk implodes, which would of course contribute to the general collapse of everything and add even more zeroes to those prices, which would never decrease even if the economy somehow survived.
My additional point is not only will the systems that play games be even less affordable than that, contributing to the general collapse of the games industry as a whole and returning it to a magazine hobby of sorts exclusively between nerds sharing blocks of code over some type of BBS, but more games are being produced with generated “placeholder” assets that the developer company “forgets to remove” before launch. It should be well-known that the best way to make sure a placeholder asset is removed before launch is to make it as obvious as possible that it’s a placeholder. Visually this means things along the lines of making objects bright pink or cubes with clashing textures or scribbled in MS Paint or putting text on it that says “PLACEHOLDER” in all caps. For audio, I guess have it be low quality or a highly copyrighted clip or song or in obvious text to speech along the lines of MacinTalk or Microsoft Sam, none of those newfangled “AI voices”. Or just have someone who’s not a voice actor read a line as flatly as possible. Then if someone actually forgets to remove it, it becomes a funny story, or maybe a small but easily resolved legal threat if it’s a highly copyrighted thing they didn’t license.
There’s also the supposed conspiracy that everything is in place to stop widespread Linux adoption from taking hold, because you can’t run Linux if you don’t have anything to run it on. Really that’s just been the case in general, between aggressive marketing to the mainstream for making sure people are most aware of Windows and Mac while also pushing smartphones and interconnected surveillance devices to the forefront, and Linux still having a higher barrier to entry, which while that’s lowered a little and adoption has slowly increased over time, it’s still a niche. Just because Valve might never be able to release the Steam Machine and Frame at this rate, and how they can’t even keep Decks in stock now for the particular market that’s buying those, that’s not the conspiracy. The real conspiracy here is any push toward cloud-only general computing to solve the problem that the cloud hosts created. Chromebooks were the warning, the modern smartphone is the weapon to ensure the future they want, making sure anyone raised on that alone doesn’t know how it really works. Ensuring Linux stays a niche is just a side effect of forcing all of local home computing into that situation. They won’t even have to outlaw it, but they’ll probably at least try.