CONCEPT
Resilience Strategy
The alternative to anticipation — deploy, observe, adapt, correct — that Wildavsky defended as the only governance strategy that historically produces safety rather than merely claiming to.
The resilience strategy is the operational complement to Wildavsky's critique of the precautionary principle. Where anticipation tries to identify and prevent harms before they occur, resilience builds the institutional capacity to absorb, detect, and correct harms as they emerge. The strategy rests on four operational principles: rapid deployment to generate information, transparent observation to detect problems, distributed feedback mechanisms to surface diverse perspectives on what counts as harm, and institutional arrangements that correct quickly when errors are identified. Applied to AI, the strategy inverts most current governance proposals, which assume anticipation is the primary tool and resilience is a backstop.
In The You On AI Field Guide
Resilience is not a passive property — it is an active institutional achievement. A resilient system has fast feedback loops, distributed decision rights, redundant capacity, and the cultural tolerance for error that allows learning to occur. Each of these must be deliberately constructed and continuously maintained. Wildavsky emphasized this construction work because the instinctive regulatory response is to build the