ChatGPT organized my thoughts:
1) Core stance
- Asymmetric risk rule: When stakes are high and the downside is catastrophic (death, long COVID, health-system collapse), bias toward precautionary measures backed by mainstream public-health evidence; relax quickly as evidence permits.
- Performance over vibes: Prefer claims that show output legitimacy—they work in practice, not just in journals.
- Institutional realism: Expertise is powerful but partial; treat it as a tool embedded in funding, incentives, and consensus machinery, not an oracle. See Turner’s The Politics of Expertise and Liberal Democracy 3.0.
2) What counts as “settled enough”
- Converging evidence: Mechanism + high-quality studies + real-world outcomes align.
- Replication or durable performance: Results hold up outside the lab.
- Transparent consensus: Commission/consensus statements publish methods, assumptions, and any minority views.
- Scope discipline: Experts flag when they leave “is” for “ought.”
3) Turner’s system critique (why experts can be right and wrong)
Modern knowledge production is a superb consensus machine built around grants, peer competition, and patent races. It rewards narrow, fundable interventions and sidelines comprehensive understanding. See Turner’s essay “Expertise in Crisis.”
4) New filter: Intervention vs. Understanding
- Intervention-science (what the system funds): fast, focused, scalable solutions (e.g., vaccines). The system executed brilliantly here.
- Understanding-science (what the system underfunds): integrated, long-horizon insight (transmission in context, heterogeneity, long-term sequelae). Expect gaps unless explicitly funded and organized.
5) Paul’s lens: don’t confuse elite signals with truth
Elite institutions can rapidly reset moral frames and police discourse. Treat elite consensus as a social signal, not evidence; always demand the data. See Darel E. Paul’s From Tolerance to Equality.
6) COVID application
- Vaccines: A genuine success of intervention-science; lifesaving and central to exit from acute crisis.
- Against revisionism: The “let it rip” / herd-immunity gambit was a harmful overcorrection that failed empirically and ethically; see Howard’s We Want Them Infected and scholarly review here.
- Policy communication: Separate “what we know,” “what we infer,” and “what we value;” publish decision trails and sunset/review points.
7) Acceptance checklist (use before taking a stand)
- Competence check: Is the expert speaking within their training, or making political/ethical tradeoffs?
- Evidence maturity: Multiple independent lines, replication, and external validity?
- Process quality: Transparent consensus with visible dissent and COI disclosures?
- Incentive stress-test: Are grantsmanship, patents, or prestige messaging outrunning the data?
- Elite-signal filter: Would you still accept this claim if it came from a low-status venue with identical evidence?
- Intervention vs. understanding: Is this a narrow fix or a comprehensive account? Adjust confidence accordingly.
8) Action rules in a crisis
- Lower-regret path: Prefer measures that reduce worst-case outcomes (vaccination, ventilation, targeted NPIs), then step down as metrics improve.
- Sunset and review: Every extraordinary measure has an expiry date, predefined rollback metrics (hospital load, excess mortality, vax coverage), and periodic public review.
- Equity check: Anticipate unequal burdens; build mitigations in from day one.
9) Red lines
- Strategies that foreseeably increase preventable harm (e.g., deliberate mass infection) fail the asymmetric-risk rule; exclude unless facts materially change and survive independent review. See Howard.
10) Update rule (pivot without whiplash)
- Predeclare the metrics that will trigger changes.
- When pivoting, state what evidence shifted, why prior guidance was reasonable at the time, and what you expect next.
11) How to write with this model
- Start with the harm baseline (what if we do nothing?).
- Place today’s claim on the evidence maturity ladder.
- Run the competence/process/incentive/elite checks and the intervention vs. understanding filter.
- State the action with sunset + metrics and list uncertainties + what would change your mind.
Key references
- Stephen P. Turner, The Politics of Expertise (Routledge)
- Stephen P. Turner, Liberal Democracy 3.0: Civil Society in an Age of Experts (SAGE)
- Stephen P. Turner, “Expertise in Crisis” (2021)
- Darel E. Paul, From Tolerance to Equality (Baylor University Press, 2019)
- Jonathan Howard, MD, We Want Them Infected (Redhawk Publications, 2023), and scholarly review here
