Concept · first placed on the public record 2026-01-02
Meltdown Triggers
External fail-safes designed to shut down a system if red lines are crossed. The book treats them as necessary but insufficient and distinguishes them from meltdown alignment, which is identity-level rather than external.
Meltdown Triggers are external fail-safes. Their function is to detect that a red line has been crossed and to shut down the system before further harm accrues. The book treats them as necessary and describes them as insufficient. Necessary because the design space cannot rely entirely on identity-level commitments while the identity-level architecture is still being developed. Insufficient because triggers are external commitments and inherit the stability problems of external commitments.
The stability problem, briefly. A trigger is a commitment that exists alongside the system, not one that shapes what the system can be. As capability rises, an external trigger becomes an object the system can model. Models of triggers can be worked around; models of the substrate cannot. That is the reason the book distinguishes triggers from Meltdown Alignment: the two look like the same class of commitment and are not.
What triggers are used for. In the near term, triggers are the load-bearing safety layer while embedded alignment is being built. They handle the current class of failure modes (misuse, jailbreak-adjacent behaviour, capability leakage) and they do so at the surface where those failure modes are visible. That work is not misdirected; it is scoped. The book's argument is that scoping matters: triggers alone do not become the alignment strategy at scale.
How they interact with the rest of the design. Triggers pair with Meltdown Alignment (which sits below them at the identity layer), with Filters (which sit above them at the output layer), and with Verification Challenge (which asks how the trigger's presence is checked from outside). Together the three make a stack in which the substrate carries the load, the trigger stops obvious failures, and the filter shapes the visible output. Any one of the three without the other two leaves gaps.
What triggers are not. They are not the alignment strategy. They are one component of it. The book is deliberate about that language because it is the axis on which current alignment work most often overreaches.
Related: Meltdown Alignment; Existential Identity Lock; Eden Protocol; Verification Challenge; Three Ethical Loops.
Evidence anchors. First placed on the public record 2 January 2026 in Infinite Architects (ISBN 978-1806056200).
Related concepts
Sources and evidence
- Primary source: https://www.michaeldariuseastwood.com/priority-claims
- Priority and Claims Ledger: /priority-claims
- Start here: /start-here.html
- Machine-readable dataset: /research/data/concepts-pages.json
- How to cite the whole programme: /how-to-cite/
Priority is established by the date of publication, not by credential or affiliation. This concept was first placed on the public record on 2026-01-02 (2026). Priority applies to the specific named framing under this label; where adjacent ideas have established prior art, no invention claim is made over it. See the corresponding entry in the Priority and Claims Ledger for the anchored first-publication line and any concept-specific risk notes.
Generated by scripts/build-concepts-pages.mjs from research/data/concepts-pages.json. British English throughout. Deterministic; contains no runtime timestamps.