Genesis from the Catastrophe Ratchet
U.S. pharmaceutical regulation (1906, 1938, 1962 Acts) was assembled one body count at a time, each expansion of regulatory reach following, never preceding, a demonstration that the previous reach was insufficient.
What decision changes?
Do not assume the catastrophe-ratchet route is available for AI: it only produces working correction infrastructure when the founding catastrophe is survivable enough that the polity persists to reform.
Most of the strongest public safety regimes were not designed; they were assembled one disaster at a time. American drug regulation is the clearest single lineage. The 1906 Pure Food and Drug Act followed muckraking journalism about contamination. The 1938 Federal Food, Drug, and Cosmetic Act — the law that first required safety testing before a drug could be sold — followed the deaths of more than one hundred people, many of them children, from Elixir Sulfanilamide, a medicine marketed without any toxicity testing at all. The 1962 amendments that created the modern clinical-trial system followed thalidomide’s birth defects. In every step, the regulator’s reach expanded after a demonstration that the previous reach was insufficient, never before.
Aviation shows the same ratchet at a different tempo: a series of near-misses failed to trigger reform, and it took the 1956 Grand Canyon mid-air collision — two airliners meeting over uncontrolled airspace, everyone aboard killed — to catalyze the creation of the FAA. Historians of regulation recognize this as the normal pattern, not the exception. Societies rarely spend political capital on hypothetical harms; they spend it on funerals.
There is something genuinely functional about this ratchet: it does produce working institutions, and the institutions it produces are anchored in vivid, specific memory of what they exist to prevent. But it has a built-in asymmetry that matters more the faster the underlying technology moves. The gap between the harm and the fix scales with how quickly the capability is deployed and how reversible the damage is. Drug regulation could afford a decades-long learning curve because each catastrophe, terrible as it was, was bounded.
That is precisely the assumption AI may break — and it is the central warning running through all eleven of these case studies. The catastrophe ratchet only works when the disaster is survivable enough that the society persists to reform afterward. Every historical case here is a case where the institution that needed fixing was still standing to be fixed. If a failure with a sufficiently capable AI system is unbounded or irreversible at civilizational scale, the most productive institution-building route in human history is unavailable exactly when it is needed most. That is not an argument for despair; it is an argument for weighting the routes that need no catastrophe — money at risk, chronic visible threat, self-refreshing memory — far above their historical share.
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?
Elixir Sulfanilamide (1938, over 100 deaths) and thalidomide (1962) each preceded, rather than followed, the regulatory capability gate meant to prevent them; aviation's FAA followed the 1956 Grand Canyon collision after earlier near-misses had not triggered reform.