Flight to the woods names the specific form of exit that senior technology practitioners began exercising in 2025 and 2026 — departing not for higher positions elsewhere but for lower cost of living, simpler lives, and distance from the transformation they could not navigate. Segal documents the pattern in The Orange Pill: senior engineers who have realized 'it's over' and moved to 'the woods' out of the perception that their livelihood would soon be gone. In Castells's framework, this is voluntary disconnection — the most dangerous response to network transformation, because the network continues to evolve in the disconnected person's absence, making reentry progressively more difficult. The motivation may be locally rational (the tool obsoletes the skill, the skill was the livelihood, therefore exit), but the consequence is structural: the network she has left will be unrecognizable by the time she considers returning.
The pattern has precedents. Every major technological transition produces a population that exits rather than adapts — craftsmen who retired rather than retrain as the factory system arrived, clerks who retired rather than learn computers, journalists who retired rather than confront the internet. In each case, individual decisions that seemed rational at the moment of departure produced collective effects that were not. The craft traditions that atrophied, the institutional memory that was lost, the specific knowledge that departed with the exits — these were costs borne by the communities left behind, not merely by the individuals who left.
What is distinctive about the AI-era flight is its pace and its concentration. The technology's rapid deployment compresses what used to be decades of transition into months, giving workers much less time to decide whether to adapt or exit. The concentration of the flight in specific domains (senior software engineering, knowledge work more broadly) produces coordination effects: when many workers in a specific niche exit simultaneously, their collective departure hollows out the mentoring infrastructure that younger workers depend on for their own adaptation.
The alternative Castells's framework recommends is not to oppose the exit impulse but to develop reconfiguration strategies — changing one's relationship to the network while remaining within it. The beaver's dam rather than the swimmer's exit. This requires institutional support that is largely absent from the current moment: structured retraining opportunities, economic cushions that permit experimentation without existential risk, and cultural recognition that mid-career adaptation is a legitimate and necessary response to technological transition rather than a personal failure.
Segal documents the pattern in The Orange Pill (2026), noting the dichotomy between those 'running for the hills' and those 'holding their ground and leaning in for the fight.'
Exit is the most common response to network transformation. Across every major transition, some significant portion of the affected population departs rather than adapts.
The costs are collective, not merely individual. The knowledge, mentoring infrastructure, and institutional memory lost to exits are costs borne by the communities left behind.
AI-era exit is unusually rapid and concentrated. The technology's pace compresses transitions into periods too short for workers to assess adaptation options; the concentration produces coordination effects in specific niches.
Reconfiguration is the alternative. Changing one's relationship to the network while remaining within it — the beaver's dam rather than the swimmer's exit — preserves the possibility of continued adaptation.