Open source has never been the sprawling community of contributors many imagine. Most essential software is maintained by a tiny core—often one or two people—doing unpaid work that companies use as infrastructure. This mismatch was manageable when contributing required effort and risk, but AI agents are erasing that friction. Mitchell Hashimoto, founder of HashiCorp, now considers closing external pull requests entirely, not from losing faith in open source but from drowning in slop PRs generated by large language models and AI agents.
Flask creator Armin Ronacher describes this as agent psychosis, where developers become addicted to the dopamine hit of agentic coding. They spin up agents that run wild through projects, producing vibe-slop: code that feels right due to statistical models but lacks context, trade-offs, and historical understanding. The quality degradation is massive, and it's getting worse.
According to SemiAnalysis, we've moved from simple chat interfaces to agentic tools that live in the terminal. Claude Code can research a codebase, execute commands, and submit pull requests autonomously. This is a productivity gain for individual developers but a nightmare for maintainers of popular repositories. The barrier to producing a plausible patch has collapsed, but the barrier to responsibly merging it has not. The best open source projects may become the hardest to contribute to.
The cost of contribution
The economic asymmetry is brutal. A developer spends 60 seconds prompting an agent to fix typos and optimize loops across a dozen files. A maintainer then spends an hour carefully reviewing those changes, verifying edge cases, and ensuring alignment with the project's long-term vision. Multiply that by a hundred contributors using AI assistants, and the result is not a better project but a burnt-out maintainer who walks away.
In the past, a developer would find a bug, fix it, and submit a pull request as a human transaction of thanks. Now that transaction is automated, replaced by a mountain of digital noise. The OCaml community recently rejected an AI-generated pull request containing over 13,000 lines of code, citing copyright concerns, lack of review resources, and long-term maintenance burden. One maintainer warned that such low-effort submissions risk bringing the pull request system to a halt.
GitHub itself is feeling the strain. As InfoWorld’s Anirban Ghoshal reported, GitHub is exploring tighter pull request controls and even UI-level deletion options because maintainers are overwhelmed by AI-generated submissions. When the host of the world's largest code forge considers a kill switch for pull requests, it's no longer a niche annoyance but a structural shift in how open source gets made.
Small open source projects are hit hardest. Nolan Lawson, author of blob-util (a JavaScript library with millions of downloads), argues that the era of low-value utility libraries is over. In the age of Claude and GPT-5, developers simply ask an AI to write a utility function instead of taking on a dependency. AI has made small libraries obsolete. If an LLM can generate the code on demand, the incentive to maintain a dedicated library vanishes.
Build it, don’t borrow it
Something deeper is lost. These libraries were educational tools where developers learned by reading others’ work. Replacing them with ephemeral AI-generated snippets trades understanding for instant answers. Ronacher’s provocation from a year ago suggests we should just build it ourselves: reduce dependencies and increase self-reliance. Use AI to help, but keep code inside your own walls. The irony is that AI reduces demand for small libraries while increasing low-quality contributions to the remaining ones.
If open source is not primarily powered by mass contribution, what happens when the contribution channel becomes hostile to maintainers? Likely a state of bifurcation. On one side, massive enterprise-backed projects like Linux or Kubernetes will have sophisticated gates and resources to ignore noise. On the other side, provincial projects run by individuals or small cores will simply stop accepting outside contributions. AI was supposed to make open source more accessible, but it has also lowered the value of contributions. When everyone can contribute, nobody’s contribution is special. The only scarce resource is human judgment required to say no.
The future of open source
Open source is not dying, but the “open” part is being redefined. We are moving from radical transparency and “anyone can contribute” to an era of radical curation. The future belongs to the few, not the many. Open source’s “community” was always partly a myth, but AI has made the myth unsustainable. The only people who matter now are the ones who actually write the code, not those who prompt a machine to do it. The era of the drive-by contributor is being replaced by an era of the verified human.
In this new world, the most successful open source projects will be the ones that are the most difficult to contribute to. They demand high levels of human effort, context, and relationship. They reject 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. We don’t need more code; we need more care for the humans who shepherd communities and create code that endures beyond a simple prompt.
Source: InfoWorld News