The Variety Trap: Why Simple Rules Fail Complex Problems

Transcript

SPEAKER_1: So last lecture we landed on correction cadence as the real strategy. But now I'm wondering — what happens when the environment is so unpredictable that even frequent corrections aren't enough? SPEAKER_2: That's exactly where W. Ross Ashby comes in. His Law of Requisite Variety says a regulator must have at least as many possible responses as the disturbances it faces. If the world throws ten problems at you and you only have three responses — you lose control. Mathematically guaranteed. SPEAKER_1: So it's not just about responding faster. It's about responding with enough range. SPEAKER_2: variety can destroy variety. Only a rich set of regulatory responses can absorb a wide range of disturbances. A narrow playbook meeting a chaotic environment — that's the Rigidity Trap. SPEAKER_1: Give me a concrete example of that trap. SPEAKER_2: Think of a manager whose only move when a project slips is to push harder. One response mode. But the disturbances hitting that team — someone out sick, scope change, tool failure — those are five or six distinct states. One response cannot absorb six disturbances. The result is burnout, or outright system failure. SPEAKER_1: Right. And Ashby uses the word 'states' specifically. What does a state actually mean here? SPEAKER_2: A state is any distinct configuration a system can be in at a given moment. For a team workflow: normal day, one person out, deadline moved up, tool down, client escalation. Each is a separate state. [short pause] The key idea is that a regulator cannot control distinctions it cannot detect or represent. If you haven't named a state, you have no response prepared for it. SPEAKER_1: So naming the states is itself part of the control system. SPEAKER_2: Precisely. And here's what feels counterintuitive — a highly optimized single routine can make you less stable when the environment shifts. Optimization often strips slack, strips variety, from your own system. You become brittle exactly when flexibility matters most. SPEAKER_1: Wait — more flexibility in the regulator creates more stability in the outcome? That feels backwards. SPEAKER_2: It does, but it holds. Biological homeostasis shows this constantly. The body maintaining stable temperature doesn't use brute force — it uses a rich set of conditional adjustments. Shiver, sweat, dilate, constrict. That variety in the regulator is what keeps the outcome stable. SPEAKER_1: Mm-hmm. And negative feedback drives those adjustments — not positive feedback. SPEAKER_2: Right. Negative feedback closes the gap between current state and target — stabilizing. Positive feedback amplifies deviation and can spiral a system out of control. A fever that keeps climbing with no correction mechanism — that's positive feedback gone wrong. SPEAKER_1: So practically — how does someone actually audit their own variety? SPEAKER_2: how many distinct states can the environment realistically enter? Two: for each state, is there a prepared response? Three: where are the gaps? For a personal schedule, map a normal day, a disruption day — say, a cancelled flight — and a recovery day. If only the normal day has a plan, variety is dangerously thin. SPEAKER_1: That's the Variety Audit. And the Triadic Response Rule follows from it — at least three contingency paths. SPEAKER_2: [emphasis] Exactly. One backup is still a fixed rule. Two options is a binary choice under pressure. Three paths force real modeling of the environment's variety. And sometimes the smarter move is to reduce the environment's complexity rather than expand your own responses. Simplify the inputs before multiplying the outputs. The operational rule: avoid approaching a complex problem with fewer response options than the environment has states that can hurt you.