ora-0031

1.18 What this chapter does not assume

This tutorial intentionally leaves several choices open.

  1. It does not require every decision target to be a model category.

  2. It does not require every decision history to be reversible.

  3. It does not assume that a learned theory has a unique presentation.

  4. It does not identify homotopy equivalence with equal numerical loss.

  5. It does not infer causal meaning from categorical universality.

  6. It does not assume that tangent lifting preserves convergence.

  7. It does not assume that every online update is an infinitesimal repair.

These boundaries are part of the framework. The chapters that follow state which special structures are used in each result and which obligations remain open.