Armin Ronacher: AI Agents Are Reshaping Software Engineering's Shared Language
Simon Willison · rss · 2026-07-15
In his essay "The Tower Keeps Rising," Armin Ronacher explores the impact of AI Agents on software engineering. He points out that a software project's "shared language" isn't just plain English or code, but rather the team's collective understanding of system concepts, boundaries, invariants, and design intent. This understanding is typically scattered across code reviews, discussions, and the process of explaining changes to others.
Before the advent of Agents, this consensus was partly maintained through collaborative "friction." For example, modifying another team's underlying modules required reading code, asking questions, and coordinating across teams. Although this slow process involved waste, its core function was to "sync human cognition"—ensuring both sides maintained a consistent understanding of how the system worked. The author worries that by eliminating friction, AI Agents might also destroy this crucial cognitive synchronization mechanism.
More from coding & agent
- Soft Clamp cuts tool-call overuse in multi-teacher distillation, from 13.7% to 9.0% — antgroup · 2026-07-21
- Agent harness memory loss and compaction are still a major usability problem — adityaag · 2026-07-21
- SpecJudge runs locally on Ollama to pick the right-sized AI model for your project — jokiruiz · 2026-07-21
- A developer maps out six design rules for CLIs that humans and AI agents can both use — yujiezha · 2026-07-21
- A coding-agent skill that forces ADHD-friendly, answer-first output — ayghri · 2026-07-21
- A set of agent skills for CAD, robotics, and hardware design — earthtojake · 2026-07-21