Hacker News
Speed at Cost of Quality: Empirical Study of Cursor AI in Open Source Projects
An arXiv paper (2511.04427) analyzing Cursor AI usage across real open source project contributions finds that AI-assisted submissions arrive faster but exhibit measurable quality degradation relative to non-AI contributions. At 52 points and 23 comments on HN, the paper provides empirical grounding that the practitioner community has been seeking for the vibe-coding quality debate. This is the first large-scale dataset study specifically examining Cursor's impact on open source contribution quality.
Source
↳ Follow the thread