Offline Fallback

For every dependency that can fail at the moment of need, create the smallest backup path and rehearse it once.

You are six miles into a backcountry hike when your phone, which holds the trail map, the weather forecast, and your ride's number, dies in the cold. The trail junctions all look the same. You realize every piece of information you need to get home safely lived inside one battery.

The move: for every dependency that can fail at the moment of need, create the smallest alternate path and rehearse it once. The cue is any digital, institutional, or external dependency whose failure would hit exactly when you need it most. Run it in three steps before the trip, not during: identify the dependency (phone navigation), build the minimal backup (a paper map and compass in the pack), and actually rehearse it once, meaning you check the map against a real junction so the first use is not the emergency. When it works, an outage becomes an inconvenience instead of a crisis, and your recovery time and losses stay inside a bound you chose in advance. The safeguard cuts the other way too: match redundancy to stakes and do not duplicate everything, because a backup for every low-stakes convenience is its own maintenance burden. Stop once acceptable recovery is covered.