Open source has never truly been a sprawling community of contributors, despite the romanticized image. Most essential software relies on a tiny core of unpaid maintainers, often just one or two individuals. That fragile balance worked when contributing required genuine effort and understanding. But AI agents are now obliterating that friction, flooding projects with automated, low-quality pull requests and pushing maintainers to the brink. This is not just a nuisance; it is a structural crisis that threatens the very open source model.
The flood of AI-generated pull requests
Mitchell Hashimoto, founder of HashiCorp and a respected figure in open source, recently considered closing external pull requests entirely. His reason: he is drowning in “slop PRs” generated by large language models and AI agents. These contributions may feel right statistically but lack the context, trade-offs, and historical understanding that human maintainers bring. Flask creator Armin Ronacher calls this phenomenon “agent psychosis”—a state where developers, addicted to the dopamine hit of automated coding, unleash agents to run wild through projects, degrading quality. The result is a massive wave of unhelpful, time-consuming submissions that force maintainers to spend hours verifying trivial changes that were generated in seconds.
GitHub is noticing the strain. As InfoWorld’s Anirban Ghoshal reported, the platform is exploring tighter pull request controls and even UI-level deletion options to help maintainers cope. If the host of the world’s largest code forge is considering a kill switch for pull requests, we are no longer dealing with a niche annoyance but a fundamental shift in how open source is made.
The brutal economics of review
The asymmetry is stark: a developer needs only 60 seconds to prompt an agent to fix typos or optimize loops across a dozen files, but a maintainer may spend an hour carefully reviewing those changes to ensure they don’t break obscure edge cases. Multiply that by a hundred contributors using personal AI assistants, and the project becomes unsustainable. The OCaml community saw a vivid example when an AI-generated pull request containing over 13,000 lines of code was rejected, with maintainers citing copyright concerns, lack of review resources, and long-term maintenance burden. One maintainer warned that such low-effort submissions could bring the entire pull request system to a halt.
The death of small utility libraries
Nolan Lawson, author of blob-util, a JavaScript library with millions of downloads, recently explored the fate of small open source projects. For a decade, blob-util thrived because it was easier to install the library than to write the utility functions yourself. But in the age of models like Claude and GPT-5, developers can simply ask an AI to generate a perfectly serviceable snippet in milliseconds. The incentive to maintain a dedicated library vanishes. Lawson argues that the era of the small, low-value utility library is over. AI has made them obsolete.
Something deeper is being lost: these libraries were educational tools. Developers learned how to solve problems by reading the work of others. When we replace those libraries with ephemeral, AI-generated snippets, we lose the teaching mentality that Lawson believes is the heart of open source. We trade understanding for instant answers.
Build it, don’t borrow it
Armin Ronacher offers a provocative alternative: just build it yourself. If pulling in a dependency means dealing with constant churn, the logical response is to retreat. He suggests a vibe shift toward fewer dependencies and more self-reliance. Use the AI to help generate code, but keep it inside your own walls. This creates a weird irony: AI reduces demand for small libraries while simultaneously increasing the volume of low-quality contributions into the libraries that remain. The very technology that was supposed to democratize open source may instead centralize it.
Two-tiered open source
This leads to a state of bifurcation. On one side are massive, enterprise-backed projects like Linux or Kubernetes. These are the cathedrals, guarded by sophisticated gates, with resources to build their own AI-filtering tools and organizational weight to ignore noise. On the other side are provincial open source projects—individual or small-core projects that simply stop accepting external contributions. The future may belong to the few, not the many. The era of radical transparency, of “anyone can contribute,” is giving way to radical curation.
The irony is profound: AI was supposed to make open source more accessible, and in a way it has. But by lowering the barrier, it has also lowered the value. When everyone can contribute, no one’s contribution is special. When code is a commodity produced by a machine, the only scarce resource is human judgment required to say no. Open source isn’t dying, but the “open” part is being redefined.
In this new world, the most successful open source projects will be those that are hardest to contribute to. They will demand high levels of human effort, context, and relationship. They will reject the slop loops and agentic psychosis in favor of slow, deliberate, and deeply personal development. The bazaar was a fun idea while it lasted, but it couldn’t survive the arrival of the robots. The future of open source is smaller, quieter, and much more exclusive. And that might be the only way it survives.
We don’t need more code; we need more care. Care for the humans who shepherd these communities and create code that will endure beyond a single prompt.
Source: InfoWorld News