“The tools allow you to move much faster if you know what you want.”
Tech
Retro gaming is being overrun with AI
Reading in data-saving mode. Read full article from The Verge website.
Published 10/23/2026 7:00:00 AM
Retro gaming is being overrun with AI
A rise in AI-powered decomps, recomps, and ports has thrust the community into chaos.
Image: Cath Virginia / The Verge
Retro video games are everywhere lately — especially in places where they aren’t supposed to be. Halo in a web browser, Wind Waker on the iPhone, Super Smash Bros. in mixed reality. Over the past month or so, these kinds of unexpected projects have gone viral with incredible frequency thanks to new AI tools, as hobbyists take apart games and then rebuild them for new platforms. But while AI has supercharged the pace of these ports, it’s also tearing apart tight-knit communities dedicated to preserving games.
Decomps and recomps — shorthand for decompile and recompile — are a form of reverse engineering that involves breaking a game down to its building blocks and then rebuilding it so it can run on hardware it was never intended for. When done manually, it’s a slow, painstaking process. A group spent years decompiling the original Animal Crossing on the GameCube, for instance. But AI coding tools have cut that timeline down significantly, and the sheer number of ports has grown dramatically in recent months.
This summer, Chris Sotraidis, an AI consultant who works on game ports as a hobby, saw a port of Command & Conquer: Generals ZeroHour — which EA made open source last year — for the iPhone and iPad. The “technical challenges” inspired him to try making a handful of similar ports, including a mobile port of The Legend of Zelda: Ocarina of Time using an existing community-built decomp of the game. Within just a few weeks, he was working on a dozen projects. He says that kind of productivity wouldn’t happen without the use of AI.
Some of those ports “would have been possible with minimal AI assistance if I’d had significantly more time,” Sotraidis says. “But I’m honest enough to say that the pace and volume would not have been possible for me without it. It would have been extremely slow by comparison. The tools allow you to move much faster if you know what you want.”
That speed has drawn many people to try out these AI tools, either for fun, the technical challenge, or for preservation reasons. It’s also a reason why there’s been what seems like a deluge of new projects popping up on platforms like Reddit, X, and GitHub. But these ports can be problematic from a preservationist perspective. One of the reasons why decomp projects exist in the first place is that users can pull a game apart and understand how it works, but that kind of understanding and knowledge is lost when AI does the work. Sure, you can put Spider-Man in a Batman game, but you probably won’t discover how certain programming decisions came to be, which can hurt the ability to restore games faithfully.
The rapid pace of AI-driven decomps has overwhelmed the DIY scene. Last week, a community-focused website that tracks decomp projects and provides various resources abruptly shut down due to AI’s impact. “It is now clear that, with the right guidance, human input, and code reviews, AI can expedite the decompilation process in a significant way,” the site’s owner, writing under the pseudonym Encounter, said when announcing the shutdown. “What is also now clear is that the community aspect of decompilation has been broken. Our projects have always been open source to encourage contributions and to benefit the community as a whole. Bad actors have been abusing this in increasingly troublesome ways.”
Those ways include taking existing decompilation projects and “finishing” them with AI, as well as racing to quickly port decomps that the user never actually contributed to in the first place. In essence, one of the big problems is people who are using AI to rush through the process while building on the work of others. This is especially an issue given the community-centric nature of the existing decomp scene. “The bottom line is that these kinds of hostile forks put little to no effort into meaningfully contributing to existing projects, tearing apart our community fabric,” Encounter wrote.
Given the ongoing challenges with video game preservation, in which it’s estimated that more than 90 percent of classic games are “critically endangered,” the rise of AI in the decomp scene can be seen as something of a double-edged sword. On the one hand, it has created a surge of interest, turning words like decomp and recomp from niche terminology to headline material, while allowing non-programmers the opportunity to experiment and test ideas. On the other hand, the speed and lack of attention to detail in many of these projects run counter to the aims of most preservationists, who try to not only keep games alive for modern audiences but also understand the way they were built. AI is a quick fix that could have long-term drawbacks.
This has created a divide, with those on the preservation side feeling disillusioned by the prevalence of slop, and those on the AI side excited by the possibilities, particularly as the tools continue to evolve. As Sotraidis explains, “I think a lot of what gets dismissed as slop today can improve as the tools get better and people learn to use them well. That doesn’t make the current criticisms invalid. Better tools aren’t automatically going to make people give proper credit or treat each other well. But I do think thoughtful creators and builders will be able to make even more amazing things, faster.”
For those vehemently against the use of AI in ports and decomps of classic games, they might be fighting a losing battle. AI tools seem to be infiltrating every aspect of game development, not just in the realm of hobbyist retro fans. Fortnite developer Epic Games recently launched a vibe-coded version of its PC games store, and it has reportedly shifted to “AI-accelerated development” for an upcoming collaboration with Disney. On X, Epic’s CEO Tim Sweeney wrote that he believes AI tools will lead to “a new golden age of gaming.”
Meanwhile, at least one major name has embraced the decomp scene. Xbox CEO Asha Sharma set up an internal division called Xbox Labs to explore the space and “turn emerging ideas into things our players and developers can experience.” She noted, “While it is hard to predict the future of gaming, it is easy to see the possibilities in making great games easier to access, giving old games new life, and allowing more players to become creators and co-creators.” (Currently, the only thing the division has teased is an experimental dashboard.)
Xbox embracing decomps is particularly notable at a time when old games are serious business. Some of the biggest releases this year are remakes of Halo: Combat Evolved and Ocarina of Time, while Resident Evil has been rejuvenated in part due to a stream of strong remakes. The question for the future is whether audiences would rather play a lavish remake of Halo or a vibe-coded port that they can take anywhere.
§
An update to Firefox’s AI browsing experience can draw from your open tabs to help you fill out forms.
An update to Firefox’s AI browsing experience can draw from your open tabs to help you fill out forms.
Mozilla is updating Smart Window — Firefox’s AI-powered browsing experience — with a new feature that can help you fill out online forms based on your open tabs. As an example, Mozilla says Smart Window can pull information from your online portfolio or LinkedIn page to paste your work history into a job application.
You can try out this feature by clicking into a form with at least four text fields. From there, you’ll see an option to “Fill from sources,” which will show all of the tabs Firefox will draw from. You can then select or deselect tabs that you want Firefox to use, and the browser will generate suggestions based on the information in them. This feature currently supports text fields on web forms, but Mozilla is working to add support for dropdown menus and PDFs. Mozilla notes that data from your source pages is shared with Google’s Gemini 3.1 Flash Lite, the AI model that generates the text suggestions.
Along with this update, you can now ask Firefox’s Smart Window to watch a certain webpage for sales, new job listings, and other updates. You can set up a task by typing “/watch” in the Smart Window chat, describing what you’re looking for, and adding the webpages you want to watch. Firefox can then check for updates on a daily or weekly basis, but only when the browser is open.
When Firefox detects the change you’re looking for, it will send you a notification through the Tasks button on the toolbar, or on your desktop (if you have desktop notifications enabled). Firefox will pause watching for updates if it doesn’t detect a change after 60 days, before stopping them completely after 90 days.
Image: Mozilla
These two features are rolling out to the Smart Window beta on desktop in the US and Canada. They’ll be available in other countries in the “coming months.”
§
The Genie 4 world model offers more realism and can create audio to accompany its interactive worlds.
The Genie 4 world model offers more realism and can create audio to accompany its interactive worlds.
Google announced today its Genie 4 world model for generating interactive experiences using AI. The company says the model has four main updates: It allows for more realistic worlds, lets users update an experience in real time with a natural language prompt, can generate audio, and uses Gemini 3.8 Flash under the hood to help with its simulations. It also can create experiences that are about two minutes long, up from one minute for the previous Genie 3 model.
Google is pitching Genie 4 as a tool that could be helpful for things like pre-visualization for filmmakers, prototyping game ideas, and making training scenarios for robots. But the company warns that, as simulations run longer, details and object permanence can shift over time.
I haven’t been able to try Genie 4 for myself — it’s launching first as a “research preview” to “trusted testers” — so I can’t tell you how it compares to the Genie 3-powered Project Genie, which I used to generate bad Nintendo knockoffs. But here are some examples from Google demonstrating what’s possible with the new model.
Genie 4 will be “coming soon” to Google AI Ultra subscribers, Google says, and the company has opened a waitlist where you can apply to be able to try it in Project Genie.
