“Scatter many seeds, some will find soil”

Salmon produce thousands of eggs with deliberate genetic and phenotypic variation. Most will die — but the variation ensures that some subset will match whatever conditions materialize. The strategy is optimal when the cost of each bet is low, the variance of outcomes is high, and prediction is impossible. Evolution has converged on bet-hedging independently in fish, insects, plants, and microbes — proof that the strategy is structurally optimal, not accidental.

Desert plants produce seeds with deliberately variable dormancy periods. Within a single crop, some seeds germinate after the first rain, some after fire, some after specific temperature sequences, and some not for years. This stochastic bet-hedging ensures that even if this year's conditions kill most seedlings, some seeds remain dormant for better years. The plant sacrifices maximum reproduction in any good year to guarantee survival across all years.

Markowitz's Modern Portfolio Theory (1952) proved mathematically that diversified portfolios dominate concentrated ones on a risk-adjusted basis. The key insight is that portfolio risk depends on correlations between assets, not just individual asset risk. A portfolio of negatively correlated assets can have lower risk than any single asset in it. This is the financial formalization of what salmon and desert plants discovered through evolution: spread your bets, and make sure they don't all fail at the same time.

VC funds invest in 20-40 companies expecting most to fail. The fund's returns follow a power law: 1-2 investments produce the majority of returns, 3-5 return modestly, and the rest return nothing. This isn't a flaw — it's the strategy. The fund succeeds at the portfolio level precisely because it accepts individual failure. The worst VC strategy is a concentrated portfolio — fewer bets means higher probability of missing the power-law winner.

Monoculture farming concentrates risk in a single genetic variety. Polyculture spreads it across multiple species and varieties. The Irish Potato Famine was a catastrophic failure of genetic concentration; traditional polyculture farmers in the same era survived because their diversified fields always had something that resisted the blight. Heirloom seed preservation maintains genetic variation as a hedge against future conditions.

Traditional drug discovery bets on one lead compound and pushes it through years of development — a concentrated strategy in a domain with 90% failure rates. Combinatorial chemistry and high- throughput screening hedge the bet: synthesize thousands of variants, screen them in parallel against targets, and advance multiple candidates simultaneously. The cost per bet is low; the variation is deliberate; success is measured at the portfolio level. This is the salmon strategy for pharmaceuticals.

Career advice typically recommends deep specialization — find your niche and go deep. But in volatile job markets where entire industries can be disrupted by AI or economic shifts, deep specialization is a concentrated bet. Skill diversification — developing competencies across multiple fields — is a bet-hedging strategy: most of your skills may lie dormant, but the one that matches the next market shift becomes your advantage.

Most cities plan for a single climate scenario (e.g., 2 degrees of warming by 2100). This is a concentrated bet on a specific future. Multi-strategy resilience planning would prepare for multiple scenarios simultaneously: infrastructure that works under 2, 3, and 4 degree warming; water systems designed for both drought and flooding; agricultural diversification for variable growing seasons. The cost of maintaining multiple strategies is real but small compared to the cost of being wrong about which future arrives.