SwiftTerm maintainer: about 80% of PRs involve AI
Understand this story
The SwiftTerm maintainer estimates that about 80% of the pull requests received so far involve AI in some form, and that this has evolved in waves. Early on, most were "vibe coding"-style superficial patches, along with Promptsplainers who used LLM explanations to override the maintainer's opinion; now AI-generated fixes have largely landed within the correct scope. This judgment comes from the maintainer's observation of contributions to their own project, and reflects the changing shape of AI participation in open-source code submissions, rather than an independent verification of code quality.
AI overview based on source reports · Updated 1 hr ago
Coverage timeline
Follow the coverage to explore different perspectives.
- Hacker News AIWhat Maintaining SwiftTerm Taught Me About AI Contributions
The maintainer of SwiftTerm, a seven-year-old open-source terminal emulator library, estimates that about 80% of the pull requests he now receives involve AI in some capacity, arriving in distinct waves from flawed "vibe-coded" patches to increasingly viable fixes. He describes the "Promptsplainer" archetype and argues that AI review tools on the maintainer's side help bridge the gap, while policies that police provenance rather than quality penalize transparency and keep good fixes in private forks. He concludes that the antidote is human accountability paired with better review tooling rather than banning the tools.
Activity for this story
Not enough continuous observations to show a trend yet.