At the start of the month, developers at Capcom—the team behind one of my favorite video game engines—gave a talk about RE Engine's future. News began to swirl about slides, statements, and translations revolving around AI. Satoshi Ishida laid the foundation for what this new "RE neXt ENGINE," aka REX, is setting out to address. Here is Eurogamer's reporting on the talk, which was in Japanese and originally reported on gamebiz.jp:

"Ishida pointed to three challenges driving the REX Project – one, games have grown so large in scope and data volume that build and iteration times are slowing development down; two, RE Engine is now used by over 2,000 developers, including international staff and people coming from other engines, requiring better multilingual and accessibility support; and three, AI needs to be built into the development pipeline rather than bolted on."

What spun reactions out of proportion were mistranslations by news outlets that took "AI" and "Generation" the unintended way, opting for generative AI instead of the "AI generation." The vocal sentiment toward generative AI in game development, at least that I observe on Bluesky and on forums, is absolutist.

What I think the bulk of the detractors fail to realize is that your games today are and were likely made with AI. The benefits of chasing down bugs, coding, and implementation are too advantageous. Jack Buser, Google's Global Director for Games, Google Cloud (who spent near eight years at PlayStation before), claimed that "nine out of 10 game developers" were using AI from a survey Google conducted at Gamescom in 2025.

"'I think what players don’t realise is that their favourite games right now were already built with AI. Those games have shipped. We did a survey around Gamescom last summer with studios all over the world. Roughly nine out of 10 game developers told us, yeah, we’re using it.

'Now you’ll see other surveys from other organisations that have that more around like 40-50%. And you might ask yourself, well, that’s still a large number. It’s still almost half of the developers out there. What’s that gap? And that gap is basically the developers willingness to tell you whether the fact of the matter is it’s being used.'"

Recent examples could be the deluge of decompiled and recompiled classic games. I see a new one pop up all the time and the news is usually accompanied with an indication of whether "slop" was used to make the PC port. There's even a new slop-derived derogatory term—"slopcomp." It's created an aggressiveness in the scene that is quite intense.

"DarioSamo pointed out that his group had already made clear its intentions to produce a port of Banjo-Tooie at some point without using AI. He also slammed the competing project over on the Decomps subreddit, calling on Banjo-Tooie Recompiled to be renamed."

I didn't realize unlicensed fan projects had rights to naming conventions, but here they are calling for changes.1 It's become so contentious, that there is a list of which decompilations and recompilations are untouched by AI—big Sneetches-with-star-bellies energy. It's not a long list either. From the introduction of said list:

"Now we’re in the mid-2020s, and AI has gotten a hold of everything, with people that have little to NO programming experience porting games completely untested, leading to a wave of buggy AI ports of games that are completely broken. And the sad part is, they overshadow the quality work of decomps/recomps done by actual people."

Just a smidge further down, the list aggregator, Mr. @funky90slion, writes the following:

"I am not against AI-Assisted Projects done responsibly as they’re different from vibecoding. HOWEVER, this list is made to showcase projects made by people who powered through any difficult tasks. And deserve to be commended for their efforts."

This distinction between "AI-Assisted Projects" and "vibecoding" is the crux of the issue.

On the same day that the Capcom news broke, Lex Friedman coincidentally published Vibe-coded apps: The vibes are good for people with specific needs. He closed out a series of wonderful tools, utilities, and apps with this:

"For decades, if the software you needed didn’t exist, you had three options: Learn to code, pay someone, or live without it. Now there’s a fourth option, and people are taking it. They’re not building the next big thing. They’re building the next small thing, for themselves. Sometimes they abandon it two weeks later. That’s allowed, too.

Vibe-coded doesn’t mean subpar. It doesn’t mean bad. And in all these cases, it means empowerment. Computers are meant to help us customize our workflows to the ways we work. LLMs make that more possible than ever."

I'm not arguing that recompiled N64 games are a utility or tool, but I will say that leveraging AI to do this process is a specific need for the person facilitating the project. It's a curiosity, a long-held desire, or a project to make playing a favorite game easier for them. I would wager it is not from some malicious, shortcut, cheating clout-chasing perspective.

It is the same fuel for creating DDDS, my app for recording modded 3DS console footage. No one was going to make that app. I was not learning how to code. I could not pay someone to make it for me. I had been living without it. With Claude Code, I was able to build the best Mac capture experience for those tens of us with modded 3DS consoles and a Mac-centric workflow.

Two days later, Friedman was back with Claude writes code; I author the apps, which I recommend you pause here and go read in full. Friedman has a deep history and background in computers, coding, and game development. It's a wonderful piece about how AI has empowered his creativity. Go ahead. It's fine, even if you get distracted and don't come back.

"As I wrote recently, not everyone’s a fan of vibe-coded apps. “Vibe-coded,” in fact, is used as an insult, the coding equivalent of “slop.” I think the term is problematic, but I understand some of the concern: Just as people are using AI tools to write crappy articles, generate crappy posters, and make crappy fake photos, there are folks using LLMs to create crappy websites, apps, and other software.

But there are good apps out there that are made at least in part with help from LLMs. In fact, I think it’s a safe bet that the vast majority of regularly-updated software you use today is made by developers leveraging AI tools in at least some ways."

I believe the same can be said for video games today. I look to the likes of Pascal Pixel and his decompilation project for the Golden Sun GBA titles—all being done by his orchestration of LLMs. He is coming from a position of passion and care for the games and turning them into readable, malleable code for players and modders alike. I don't think anyone could look at his work and posts and call his authenticity into question. Back to Friedman:

"I know of zero “vibe-coded” apps that started with a single prompt and then were immediately released. There’s a reason we’re not flooded with more amazing apps than ever before. Great apps take time. There is a real problem of too many vibe-coded apps hitting the App Store, lacking in craft, skill, and humanity.

The question, to me, isn’t how much code was written by a human. It’s how much the app was authored by a human, which I think is a more nuanced question. How much care, thought, and humanity went into the app matters far more than tabs vs. spaces or a manually-vs.-machine-built HStack.

Movie editing and newspaper layout aren’t less impressive now that we use software vs. scissors. WYSIWYG text editors don’t make the written word less compelling than typewriters...

...The person using the tool has direct influence over the quality of the final product."

AI is and has been a tool. The same rules that apply to a hammer and chisel apply to a stack of GPUs burning through tokens. The tool doesn't make the art, the product, the application—the human does with the tool. When you hit the publish button, that work is yours, not the LLM's. If that recomp is a buggy mess and it's easier and more delightful to just emulate the old-fashioned way, blame the developer, not the methodology of development. But if that recomp is sound, sturdy, and delightful, I say praise the person and how they use the tools of the trade. Same goes for modern releases and the like. It is like I wrote in Is Coding Not Art?:

"When it comes to AI in game development, I make the call when I see it. Does the game look good, play great, and resonate? The devs used the tool the right way. If it looks sloppy, I say that's the developer's fault, not the tool's."

If there is any studio I trust to build and upgrade their engine for the "AI generation," it is Capcom. If I was ever looking for a reason to play Golden Sun on my computer, I'd look at Pascal's project. I'm looking for the games and projects authored by humans.

Footnotes

  1. Seems beyond silly to be this upset about a project labeled as an "alpha" as well. ↩