The sev-2 is paging (#172’s framing, fifteen days on, grimly upgraded). Italy locked down towns TODAY — the outbreak has a European beachhead with community transmission and a case-fatality signal nobody can yet denominate (testing coverage IS the denominator problem; you can’t compute a rate when you’re only sampling the sickest — #144’s survivorship-biased floor, now epidemiological). South Korea’s numbers are climbing on a church-cluster superspreading event (#172’s dispersion lesson, empirical within a fortnight). The Diamond Princess cruise ship — quarantined in Yokohama with infection spreading INSIDE the quarantine — has become the world’s grimmest natural experiment in closed-system transmission. And Mobile World Congress, a 100,000-person conference, was simply CANCELLED — the first major domino of what I now suspect is a very long row.
At work I did the staff-engineer version of believing my own dashboard: wrote an RFC titled “Remote-First Contingency” and scheduled a WFH FIRE DRILL for next week — the whole company, one full day, everyone remote, find what breaks. The exec response ranged from supportive to gently amused (“a bit much?”). Maybe! The #135 doctrine says rehearsal is the only rollback for irreversible events, and “the whole company can’t enter the building” is about as irreversible as scenarios get. If it’s overcaution, we lose a Tuesday. Asymmetric bet. Taking it.
Markets, notably, spent the fortnight at all-time highs, pricing containment with a confidence the epidemiology does not obviously support. One of these two dashboards is wrong. (#136: all valuations are load-bearing beliefs. Belief uptime: under observation.)
TIL: the fire-drill findings, pre-registered for next entry — my prediction: VPN capacity fails first, meeting culture fails second, and the real gap will be something nobody listed. Grading in fifteen days, hopefully with the luxury of it still being hypothetical.