summary - Originally written about RHIC-era “killer strangelet” concerns (submitted 2000, four years pre-publication), later cited in LHC-era discussions; argues that even a rigorously-derived small probability bound is not actual reassurance once priced as an expected-cost calculation against a global-catastrophe payoff, and that the collider-physics community’s implicit risk tolerance for tiny-probability catastrophic outcomes is far more permissive than established radiation-safety policy elsewhere — estimating a consistent policy would require risk bounds below roughly 10^-15/year.
relevance_note - The earliest systematic version of this slice’s central meta-question (is a small stated probability actually reassuring, and is it being judged by a consistent standard); predates and foreshadows Ord et al. (2010) and the Tegmark-Bostrom anthropic point.