Completion is a Substrate, not a UI
Summary
This article argues that Incremental Completing Read (ICR) is a structural change in how interface cost scales with data, not a mere UI nicety. It explores the ubiquity and composability of ICR, the cognitive benefits of cued recall and recognition, and how Emacs exposes completion as a programmable substrate to enable modular, extensible toolchains. It also discusses data organization shifts from hierarchical to flat structures enabled by ICR, with examples and a roadmap for practical implementation.