Guided questions

Because agricultural claims are being converted from assertions into audited assertions, and that conversion is happening from several directions at once: disclosure regimes, lending conditions and supply-chain commitments.

Because the asset is the evidence model, not the software. A model that is not widely instantiated is worth nothing, and price is not the barrier to instantiating it — integration effort is.

By pricing reliance. When an institution depends on an assertion it did not produce — for a disclosure, a loan condition, a supply-chain commitment — that dependence is the billable event.

Copying the software is easy. Occupying the position is not, because the position is a shared representation that other people have already implemented.

The science we depend on is not new measurement technique — it is the honest composition of evidence produced by different methods with different uncertainty structures.

Partly, and the parts are separable. The evidence model and its conformance suite are built and can be run against by anyone. The trust kernel is built and has supported one external reliance path in a pilot.

Pricing reliance. Unassisted adoption. Institutional acceptance of technical reliance as legal reliance. Standardisation by a body we do not control.

Three findings would be fatal rather than inconvenient. First: institutions satisfy their obligations with cheap attestation and never demand verifiable evidence. Second: no institution will pay for reliance, at any price, because the budget line does not exist. Third: an incumbent with distribution imposes an adequate model before ours is adopted outside the portfolio.

Wave 0 is one programme with four simultaneous readings. For investment it tests willingness to pay for reliance. For technology it puts the model in front of integrators we do not control. For science it compares stated to observed uncertainty on real multi-method data. For diligence it attaches a pre-registered success criterion to each unresolved claim.