Rhythm and Dwelling

In the AI Story

Sennett's analysis of rhythm drew on neuroscience research showing that motor skill development depends not merely on practice volume but on practice structure. Sleep-dependent memory consolidation studies demonstrate that motor skills improve during rest periods following practice—the musician who practices, sleeps, and practices again often finds the passage easier the second day not because of additional practice but because of the neural consolidation that occurred overnight. The intervals are not dead time; they are active developmental periods during which the brain reorganizes what was learned, discards unsuccessful strategies, and strengthens the neural pathways that produced successful performance. When practice is continuous—when iterations follow immediately one after another without rest—this consolidation is incomplete, and the developmental benefit of the practice is reduced.

The rhythm of AI-assisted work is fundamentally different from the rhythm of traditional craft. The developer working with Claude Code may complete thirty cycles of description-generation-evaluation in an hour—a velocity unimaginable in hand-coding. Each cycle provides feedback and opportunity for refinement, which is the structure of learning. But the cycles are so fast, and the intervals between them so short, that the dwelling Sennett identified as essential to deepening may not occur. The practitioner is always in the next iteration before the previous one has been fully metabolized. The work can be extraordinarily productive—Sennett would not deny this—but productivity and developmental depth are not the same thing, and a practice optimized for one may undermine the other. The question his framework raises is whether organizations deploying AI tools will structure work to include the dwelling time—the protected pauses, the deliberate intervals, the slower rhythms—that allow practitioners to consolidate what they are learning, or whether the economic pressure to maximize throughput will eliminate the temporal architecture that deep learning requires.

Origin

The concept of dwelling comes from Heidegger's ontology—wohnen, the mode of being-in-the-world that is more fundamental than building—but Sennett adapted it from Heideggerian abstraction into concrete ethnographic observation. He documented the way craftspeople spoke about the importance of 'letting the work sit'—the furniture maker who leaves a piece unfinished overnight and returns to find that problems visible the next day were invisible in the heat of production, the programmer who walks away from a stubborn bug and returns after a weekend to see the solution immediately. The dwelling is not laziness or procrastination but a necessary phase in the rhythm of creative work—the fallow period during which unconscious processing clarifies what conscious effort could not resolve. AI's continuous availability threatens this phase by making the next iteration always available, converting what would have been a necessary pause into an opportunity for additional work.