Could AI build a real alternative to Windows? A wave of “vibe-coded” operating systems — OSes written mostly or entirely with AI coding assistants — suggests the answer might be closer than you’d think. Here are three wild examples from a new How-To Geek roundup.
Vibe-coded OS (Credits: How-To Geek)
What Is a Vibe-Coded OS?
The term “vibe coding” means describing what you want to an AI coding assistant — like Claude Code — and letting it generate the software. Hobbyists are now applying that to the most complex software project imaginable: a full operating system, kernel and all.
3 Vibe-Coded Operating Systems Worth Knowing
1. VibeOS: This hobby OS runs on ARM64 CPUs, ships its own kernel and a graphical desktop, and was built from scratch across at least 64 Claude Code sessions, according to its GitHub page. You can download it today and run it on a QEMU virtual machine — or even on a Raspberry Pi Zero 2W. It even runs DOOM, which has become something of a rite of passage for vibe-coded OS projects.
2. SlopOS: The name is a joke, but the OS is real — it boots on bare metal, no virtual machine. Its three “kernel wizards” wrote their own Intel display driver from scratch, along with their own TCP networking stack. It also ships a “Wheel of Fate” feature that makes every boot a gamble — will it even start?
3. The other VibeOS: A second project shares the name, and the How-To Geek author notes that unoriginal naming in this scene is almost poetic.
Could a Vibe-Coded OS Replace Windows?
Honestly: not yet, and probably not soon. As the article notes, “not everything works, some stuff is not even tested.” These are hobby projects, built by individuals with AI assistance rather than the teams of professionals who built Windows, macOS or Linux.
But the trajectory is remarkable. It took vast sums of money and thousands of professional hours to create an OS that could run DOOM the first time around — and now one person with an AI assistant can pull off something comparable from their desk. The bigger story is what that says about where software creation is heading: regular people using AI to build tools that once required entire companies.
Also Read: Y Combinator’s Garry Tan Spends Up to $200K a Year on AI Agent Tokens
Source: How-To Geek — read the full story there.
