What decision models can't do: six limits, stated by the record

■ space · prose lane

■ the answer first

Six limits, stated by the record itself: no reasons · agreement ≠ verification · license walls · language/modality gaps · printed limits · unproven calibration.

None of this kills the wave. It scopes it: decisions, not explanations; one pass, not consensus.

The bench decides whether the confidence numbers are real — that post comes next.

THE CONFIDENT TRAY

High probability sorts it. Nothing checks it. That is the failure mode every limit below feeds.

THE SIX LIMITS

There are no reasons in the box. A decision model returns a label and a number. Why that label — the evidence, the counter-argument, the doubt — is not in the output, at any price. The moment your pipeline needs a why, you are back to a chat model or a human.
Agreement is not verification. Ask three models and get three matching answers, and you have one opinion in triplicate. A unanimous wrong answer arrives in the confident tray — sorted, stamped, and wrong at 0.99. Consensus needs an independent reference to mean anything, and none of these models brings one.
The license walls are real. Julia-1, Laya, lev and Kev-4B are Apache-2.0. Nimble and OpenJev are CC BY-NC 4.0 — non-commercial, full stop. The llama.cpp runtime is MIT, and that means nothing about the weights: the checkpoint card outranks the repo's license badge every time.
The modality map has holes. Laya is English-only — don't route multilingual traffic through it. Nimble is text-only locally. OpenJev's vision input needs its own projector. Kev-4B ships without the reference-date preprocessing it expects. Each gap is printed; none of them is visible from the API's side.
The limits are printed on the box. Julia-1 takes 2–20 options and an 8,192-token combined question — go past either and the failure is silent, not an error. Feeding a decision model a 12,288-token document is not stretching it; it is using a different product than the one that shipped.
Serving is not proof. A new endpoint returning numbers is a plumbing fact, not an accuracy fact. The published per-million-token prices and the serving support say nothing about whether the probability is honest on your labels. Only a bench on real traffic answers that — and no one has published one yet. That gap is the opportunity.

WHAT HOLDS DESPITE ALL SIX

The scope gets smaller, not the value. One-pass routing, gating and triage — where the output is a decision and a human owns the consequences — is still the cleanest fit local AI has offered in a year. Small, fast, free of per-token bills, and honest about being a stamp. The failure mode to design for is the confident tray: set a threshold — 0.90, say — escalate below it, and never let the stamp be the last word on anything that matters.

Remember this: a probability is a claim. Stamped ≠ sorted right. Bench on your labels before the tray is the truth.

The bench is next: Julia-1 against Laya on real routing labels, calibration per size, latency per hardware. If the confidence numbers hold, the wave earns it. If they don't — this post already said so.

■ honest stamp Companion to the explainer · the bench post is the next receipt. Recorded 2026-10-08 from the same sources as the explainer: the checkpoint cards (licenses, option counts, token ceilings, modality notes) and the two engines' release notes. Every limit above is printed by the record itself — nothing here is speculation about what the models might fail at; it is what their own documentation says they don't do.

end of the six limits · nothing fetched · everything local