Workshop: Maintaining Project Quality and Growth With AI-G...
summary
Maintainers discuss concrete policies and tradeoffs: Fedora requires an assisted-by tag and human final signoff; projects try tooling like CodeRabbit reviewer which catches correctness but not idioms; one project crosses 1,000 contributors and worries about scaling reviewer education; Crossplane accepts a large AI-generated feature worked by placing it in a SIG repo even though no one fully understands it; attendees debate tagging model versions, levels of AI assistance, and reviewer workload.
AI-written summary · about
chapters (11)
- 0:00Introduction
- 1:06AI policies for open source
- 4:47Trust in AI contributions
- 9:44Effective AI policies
- 14:00AI slop and reviewer workload
- 17:53Transparency in AI contributions
- 22:10Maintainer review processes
- 26:50AI assistance vs contributor learning
- 30:18Legal implications of AI code
- 34:13Educating contributors with AI
- 35:59Closing remarks
related talks
34mUnconference Session: How Should Maintainers Navigate (and Review) AI-based PRs?
46mPanel Discussion: AI and Community: Shifting Roles, Rising Impact
31mAI-Assisted Contributions and Maintainer Load
43mWriting for Machines: How To Capture Your Project's "Vibe" and Survive AI Slop
01h 01m"Looks Good to Me": A Practical Guide to Handling AI-Generated Code
21mHow To Make AI-Assisted OSS Contributions Review-Ready
50mPanel: Beyond The Vibe: Code Quality First
46mPanel Discussion: AI's Impact on Developers