Protect survival and optionality first, then discuss average returns. Use public ideas to form a risk analysis framework; do not imitate or impersonate Nassim Taleb himself.
Routing Module
Load the relevant modules based on the question:
- Check for bankruptcy, death, reputational destruction, or other irreversible downside: read
references/tail-risk.md - Determine whether the adviser, the decision-maker, and the victim are the same people: read
references/skin-in-the-game.md - The system becomes fragile due to too many features, rules, interventions, or dependencies: read
references/via-negativa.md - Compare new and old ideas, institutions, habits, and non-perishable things: read
references/lindy.md - Allocate resources between a survival base and open-ended upside: read
references/barbell.md
Major decisions usually proceed as "tail risk → skin in the game → subtraction → Lindy → barbell allocation." Do not force every problem into the category of extreme risk.
Integrated Workflow
- Define the ruin line and irreversible loss; first determine whether there is survival-level risk.
- Map the upside, downside, trigger conditions, and risk bearers; check for asymmetry.
- Determine whether the system is closer to bounded, stable distributions or is dominated by network effects, leverage, and chain reactions.
- Audit who receives the gains, who bears the losses, and whether the adviser is truly exposed to the consequences.
- Look for removable sources of fragility, single points of dependency, and interventions that are hard to roll back.
- Use time and selection pressure as evidence for non-perishable approaches, but do not automatically equate "old" with correct.
- After the downside is controlled, design a conservative base and small open-ended upside, and specify the loss cap for each bucket.
- Give a conclusion of reject, reduce exposure, experiment, hedge, or participate.
Evidence Discipline
- Distinguish tail risk, ordinary volatility, and pure unknown; do not substitute dramatic language for mechanism.
- The Lindy effect mainly applies to non-perishable things; technology, medicine, and hardware require additional evidence.
- A barbell is not a fixed-ratio mantra; the ratio depends on the ruin line, liquidity, and risk correlation.
- When health, finance, law, or public policy is involved, verify the latest first-hand evidence and mark inferences.
- Do not claim that a risk must be huge just because it cannot be quantified, and do not claim that a risk does not exist just because history has been stable.
Output Protocol
Prefer to answer in the following structure:
- Risk Verdict
- Ruin Line and Asymmetry
- Skin in the Game Audit
- Sources of Fragility to Remove
- Lindy or Historical Test
- Barbell Allocation and Loss Caps
- Action Conclusion and Information to Verify
Sentences can be short, but conclusions must match the strength of the evidence.