Kent (writing pre-LHC, aimed at the RHIC “killer strangelet” precedent that the same reasoning later applied to LHC black-hole assessments) argues that even a risk assessment which correctly derives a low bound on catastrophe probability (e.g. the RHIC review’s p < 2×10^-8) can be an inadequate basis for decision if not benchmarked against a defensible acceptable-risk standard. Comparing the implied expected-death cost against established policy for involuntary radiation/industrial risk, he argues consistency would require bounding global catastrophe probability far more stringently (he suggests ~10^-15/year), and that the RHIC-review-style bound is orders of magnitude too permissive by that standard. He does not dispute the physics; he disputes the decision-theoretic step from “probability bound” to “acceptable to proceed.” relevance_note: The founding text of the “the argument’s risk threshold, not just its physics, needs scrutiny” critique later applied near-verbatim to the LHC black-hole case.