Summary

Armin Ronacher reflects on maintaining Pi's open source project in the age of AI-generated contributions, describing how LLM-assisted issue reports and pull requests are degrading the quality of open source collaboration. He details how 'slop issues'—bug reports run through LLMs that add confident but wrong diagnoses—actively mislead both human maintainers and AI agents trying to fix them. He critiques how LLMs systematically over-engineer fixes by adding tolerant readers and fallbacks instead of enforcing invariants, growing needless complexity. The post describes Pi's own workflow of using AI agents to triage issues while explicitly instructing them not to trust LLM-generated analysis. Ronacher argues that AI has increased the volume of code and projects without increasing the number of maintainers or users who need software, fragmenting effort that should be shared. He concludes that open source needs more human-to-human collaboration, not more isolated work with machines.

Key Insight

AI-generated contributions to open source are degrading collaboration by increasing volume without increasing quality, replacing human-to-human communication with isolated human-machine workarounds, and systematically favoring local complexity over global correctness.

Spicy Quotes (click to share)

• 4

  A bad issue that contains a plausible but wrong diagnosis creates extra work.

• 9

  If you do not know the root cause, say that. I too can operate a clanker, and I would rather do this myself than use your slop.

• 7

  Almost always, the correct fix is not to handle the bad state, but to make the bad state impossible.

• 5

  All these models see a local failure and try to locally defend against it. As maintainers we have to keep pulling the conversation back to the global invariant, which is harder than it should be, and it's laborious.

• 9

  If your clanker shits on someone else's issue tracker then it's not the fault of GitHub, it's yours alone.

• 8

  AI has not increased the number of people who need software, or the number of maintainers who can review it. It has mostly increased the amount of code and the number of projects competing for attention.

• 6

  We need stronger foundations, not weaker ones. Open Source needs more collaboration, not more isolated work with a machine.

• 7

  Agency lies with humans, not with machines. Calling these things agents I still believe is a mistake, but alas.

Tone

critical, opinionated, reflective