Failure: Reform Decay
Glass-Steagall era banking constraints, built from Depression-era catastrophe, eroded on roughly the timescale over which the generation that lived through the founding catastrophe left the relevant institutions — culminating in repeal in 1999 and a reproduced failure in 2008.
What decision changes?
Treat correction-channel integrity as time-dependent, not a standing property once certified: build in an explicit decay term or a mandatory re-certification trigger tied to elapsed time since the last exercised correction.
A safeguard maintained by memory of a specific catastrophe decays on roughly the timescale of a human working generation — unless something re-derives its justification independently of that memory. American banking regulation after the Great Depression shows the full arc with unusual clarity, because both endpoints are documented catastrophes.
The 1930s bank runs and speculative collapse produced Glass-Steagall’s separation of commercial and investment banking, federal deposit insurance, and the Securities and Exchange Commission. These held for decades. Then, as the generation that had lived through the founding catastrophe retired out of the banks, the regulator, and Congress, the constraints eroded — not in one dramatic repeal but piece by piece: supervisory reinterpretations here, exemptions there, each individually defensible, each met by the reassuring observation that nothing bad had happened. By 1999 the formal repeal of Glass-Steagall was almost an afterthought; the wall had been hollowed for years. Nine years later, the 2008 crisis reproduced, in modernized form, several of the exact failure modes the original constraints had targeted — with risk migrating into off-balance-sheet vehicles the supervisory system could not see, and credit ratings paid for by the very issuers being rated.
The intervals are worth stating plainly, because they suggest something like a decay constant. From founding catastrophe to repeal: roughly six decades — two working generations. From repeal to reproduced catastrophe: roughly one. The erosion was not driven by villains, mostly; it was driven by the ordinary epistemics of forgetting. Every year a rule against a rare disaster goes untriggered, the evidence that it is unnecessary accumulates and the evidence that it is essential does not. A safeguard that works looks, from inside, exactly like a safeguard that was never needed.
The implication for AI governance cuts against how oversight is usually imagined. A safety regime certified once — an eval passed, an audit cleared, a framework adopted — is treated as a standing property, when history says it is a decaying one. Correction capacity needs an explicit expiry: mandatory re-certification triggers, forced exercises of the real handles, and mechanisms that re-derive the case for each constraint for people who never saw the failure it prevents. This is the mirror image of memory refresh (the Venice card): reform decay is simply what happens where refresh is absent.
One of eleven historical case studies in Institutional Genesis, Memory, and Decay — see the overview for the full life-cycle map, or read the complete appendix.
What would count as evidence?
The interval from the 1999 Glass-Steagall repeal to the 2008 crisis was roughly a decade; the interval from the original 1930s catastrophe to the repeal was roughly six decades — with risk migrating off the regulator's checked balance-sheet abstraction into off-balance-sheet vehicles the correction system could not see, compounded by issuer-paid credit rating agencies.