1 November 2026

Essayist argues AI-written code and docs erode trust in software

• Kyle Kingsbury argues that using large language models (AI systems that generate text and code) weakens trust in documentation and software.
• His point is that polished signs of a mature project are now easy to fake, so engineers may grow suspicious of new projects.
• The core open question is how maintainers, the people who look after a piece of software, can prove to users that their work is high quality.

How it was covered

Simon WillisonDaily notes and links

A quotation from Kyle Kingsbury's Jepsen ParticleDB results, on LLMs and software trust