Getting Lost (Solnit)

In the AI Story

Applied to the AI moment, Solnit's framework reveals that the tools' most valuable function may not be the one the industry celebrates. The industry celebrates efficiency—converting questions into answers, problems into solutions, intentions into artifacts at unprecedented speed. Efficiency operates within existing frameworks, optimizing within existing boundaries. It does not create new frameworks or redraw boundaries, because creating and redrawing require the willingness to abandon the existing map. The AI tool's capacity to surprise—to make an unexpected connection, to link domains the builder had not considered, to produce an output that reframes the question—is the capacity for productive disorientation. But this capacity is crowded out when the tools are used purely for efficiency, when every prompt seeks a predetermined answer, when the builder never allows herself to be lost because the tool is always there to provide the next step.

Solnit writes about walking as a practice of productive disorientation—moving at a pace that allows the mind to wander, to encounter the unexpected, to get lost in the specific way that precedes discovery. The practice is physical, but the principle is cognitive. The mind that moves at the speed of instant query-response never gets lost, because the tool provides the next answer before the question has fully formed. The mind that moves at the speed of embodied attention—slow enough to notice, slow enough to be surprised—can encounter what it did not expect. The Berkeley researchers documented task seepage, the colonization of pauses by AI-accelerated work. Those pauses were not empty—they were the cognitive wandering through which unexpected connections form. When the pauses disappear, the wandering disappears, and the mind becomes efficient and incurious.

The practice Solnit prescribes is not refusal of AI tools but the deliberate cultivation of conditions under which productive disorientation remains possible: asking questions you do not know the answer to, allowing yourself to be surprised by outputs rather than evaluating them against predetermined standards, building pauses into workflows where the conscious mind releases its grip and the associative, the unexpected, the genuinely new can surface. The twelve-year-old asking "What am I for?" is lost—she does not know the answer, cannot predict where the question will lead. Solnit's framework insists this gap is not a problem to solve but a condition to inhabit, and that the answer will arrive not through analysis but through the willingness to sit with not-knowing until something genuine emerges.

Origin

A Field Guide to Getting Lost emerged from Solnit's practice of walking in the American West—deserts, mountains, the specific landscapes where disorientation is not merely metaphorical but actual, where the person who loses the trail must navigate by attention to detail, by improvisation, by the willingness to be changed by what the landscape teaches. The book is structured as a series of essays that refuse to resolve into a unified argument, each exploring a different dimension of lostness: geographical, emotional, historical, metaphysical. The refusal of unity is itself the argument—that coherence is sometimes the enemy of truth, and that the willingness to remain in productive confusion is a discipline worth cultivating.

The concept builds on earlier traditions: the Romantic valorization of wandering (Wordsworth, Thoreau), the phenomenology of disorientation (Heidegger's un-handedness, Merleau-Ponty's chiasm), and the feminist epistemology of situated knowledge (Haraway). Solnit's synthesis is distinctive in making lostness a practice rather than an accident—something to cultivate deliberately rather than something that merely happens when competence fails.

Key Ideas

The Person Who Knows Cannot Discover. Discovery requires encountering something outside the existing framework. The person who always operates within a known map can only find what the map already contains. Getting lost is the precondition for finding what the map did not predict.

Two Kinds of Not-Knowing. The not-knowing that produces paralysis ("I don't know what will happen, therefore I won't act") versus the not-knowing that produces exploration ("I don't know what will happen, therefore what I do might matter enormously"). The distinction is emotional before intellectual, living in the body's response to uncertainty.

Productive Disorientation Requires Pauses. The mind moving at screen-speed—instant query, instant response—never gets lost because the tool eliminates the gap where disorientation lives. The pauses AI tools colonize through task seepage are the pauses where the wandering mind makes unexpected connections.

Walking as Method. Solnit's literal practice of walking at pedestrian pace through landscapes models the cognitive practice of moving slowly enough to be surprised, to notice what a faster pace would miss, to allow the unexpected to register before the framework assimilates it.

The Twelve-Year-Old's Lostness. The child asking "What am I for?" inhabits genuine not-knowing. The gap between the question and any answer is not a problem requiring immediate resolution but a condition permitting growth. The adult's job is not to eliminate the uncertainty with a reassuring answer but to help the child develop the capacity to sit with it productively.