Core Principles

LDD rests on six principles. They're listed here with one-line summaries; each has its own page with the full argument and (eventually) examples, anti-patterns, and case studies.

1. The Learning Loop is the Atom. The smallest unit of work is a full frame → slice → build → validate → decide loop.
2. Unvalidated Decisions Are Toxic Inventory. Decisions you haven't tested accumulate like debt.
3. Context Engineering Beats Prompt Engineering. Good context + sloppy prompt beats bad context + perfect prompt.
4. Learning Loops Need Orchestration, Not Management. Someone must design the experiment and interpret the results. That's a function, not a job title.
5. Roles Consolidate, Not Disappear. AI doesn't eliminate roles; it blurs their boundaries. Same boat, different lenses.
6. The Learning Cadence Sets the Pace. Loop length emerges from how fast you can close a learning loop, not from a calendar.