Defining AI-Native Systems: Core is Revision Authority, Not AI Count

tianyin_xu · x · 2026-08-09

ACM SIGOPS published an article by Cheng Tan titled "Defining AI-Native Systems: Autonomy as Revision Authority."

The piece proposes a core insight: AI-nativeness is not determined by how much AI a system contains, but rather by who is allowed to change the system. A full writeup is available on arXiv.

Original post →

More from Research

Research channel →