Open Questions

As this approach matures, we'll need to answer:

1. How do you maintain architectural coherence while running dozens of thin-slice learning loops? (Cockburn's hexagonal architecture + TDD help, but this is an open challenge.)
2. How do you prevent "learning loop whiplash"? Constantly validating, constantly changing direction, never shipping anything complete.
3. How do you measure learning in a way that's quantitative enough to track but doesn't turn into vanity metrics?
4. How do you teach people to think in learning loops when they've spent years thinking in sprints? (This applies to teams and to solo builders who've internalized "just ship it" without "just validate it.")

These are real problems. We'll solve them as we practice.