Scope and the Correction-Capacity Assumption
The framework applies only while civilization still has enough capacity to notice, evaluate, and constrain frontier systems — this is a scope condition, not a guarantee.
What decision changes?
Before applying any tool from this book, ask whether society can still notice relevant changes, interpret evidence, contest it, constrain deployment, and recover from a failed correction attempt.
Every alignment theory begins somewhere. This one begins after assuming a hopeful answer to a narrow question: can humans still notice, understand, refuse, and redirect what frontier systems are doing? If the answer is no, the rest of the framework may be correct and still arrive too late.
The book names five abilities this capacity requires: noticing relevant changes soon enough, interpreting what the evidence means, contesting it under pressure from labs and markets, constraining deployment on the evidence, and recovering if a correction attempt fails. The threshold itself is unknown and treated as a scope condition, not a measured constant.
The chapter’s central counterexample is “alignment after capture”: a world with visible labs, active regulators, and articulate safety rhetoric where the real selection handles have already shifted — procurement depends on frontier systems, regulators depend on model-generated analyses they cannot independently check. The institutions look more articulate while becoming less causally effective. Boundary audits, correction-channel dashboards, and successor-certification committees can all keep producing reports in a world where none of them still change what gets deployed.