
[YOU] on AI opens with an image of swimming inside a fishbowl—the set of assumptions so familiar you have stopped noticing them, the water you breathe. Borgmann’s analysis provides the philosophical scaffolding for that image. The fishbowl is a technological environment, and its walls are made not of glass but of device-mediated habits, expectations, and evaluative frameworks the inhabitant has internalized so completely that they feel like the shape of reality itself. The book’s geological metaphor—each hour of debugging depositing a thin stratum of comprehension, accumulating over years into something the senior engineer can stand on—is Borgmann’s internal goods account translated into engineering phenomenology: the focal practice of wrestling with code that will not work deposits the geology; the AI device delivers the working code without the deposition.
The cycle’s account of Segal writing with Claude—the moment he almost kept a passage that was eloquent but not his, then deleted it and spent two hours handwriting until he found a version that was—is a Borgmann moment: the recognition that the server model of creative work delivers the commodity while the hearth model produces the understanding. The deleted passage was a device output. The handwritten version was a focal practice. The difference is not visible in the final text. It is entirely visible in the practitioner.
The Berkeley study the cycle cites—AI tools in a 200-person tech company, eight months of embedded observation, finding that workers took on more tasks, expanded domains, and filled previously protected pauses with additional productive activity—is exactly what Borgmann’s framework predicts. The commodity is being produced in greater volume. The engagement has shifted from depth to breadth. The study measures output. It cannot measure whether the output was produced through focal engagement or through device-mediated production—the only distinction that matters for whether the practitioner is being deepened or thinned.
Borgmann’s most important contribution to the cycle is the concept of disburdening as a two-edged operation. The elimination of obstructive friction—dependency management, configuration files, the mechanical connective tissue the Trivandrum engineers were glad to lose—is an unambiguous gain. The simultaneous elimination of productive friction—the ten minutes of unexpected system behavior embedded in the four-hour configuration block, the moments when something went wrong in ways that forced understanding—is the hidden cost the device paradigm structurally conceals. Borgmann gives the cycle the vocabulary to distinguish the two when the device absorbs them indiscriminately.
Born in Freiburg im Breisgau in 1937, Borgmann studied philosophy in Germany and the United States before joining the faculty at the University of Montana in Missoula, where he spent his entire career. His intellectual formation was shaped by the German phenomenological tradition—Heidegger, in particular—but his project was always more empirical and less apocalyptic than his sources: an effort to understand the specific, practical ways technology shapes human life, rather than a grand narrative of technological fall or salvation.
Technology and the Character of Contemporary Life (1984) introduced the device paradigm with sufficient rigor and range of examples to found a new subfield of philosophy of technology. Crossing the Postmodern Divide (1992) extended the analysis to cultural theory. Holding On to Reality: The Nature of Information at the Turn of the Millennium (1999) applied the framework to the information revolution. His final published work addressed AI and robotics in a lecture at the Garrett-Evangelical Theological Seminary—a manuscript titled “Artificial Intelligence and Robotics: Promise and Peril” housed in the Mansfield Library at the University of Montana.
Borgmann was recognized by a handful of contemporary philosophers as having produced the most precise existing instrument for technology criticism—not the apocalyptic denunciation of a Heidegger or Ellul, but a structural analysis that could distinguish between what technology delivers and what it quietly removes. L.M. Sacasas, whom Borgmann himself endorsed as a thinker whose explorations might transform a culture urgently in need of transformation, extended the device paradigm into the AI age with his three-stage framework: mechanization, automation, and animation—the replacement of human judgment, creativity, and initiative with machine capability.
The device paradigm. Every technology can be analyzed by separating the commodity it delivers from the engagement it requires of the user. A device is characterized by the concealment of its machinery and the disburdening of the user from the skill, effort, and attention that the pre-device version of the activity demanded. The commodity remains available. The engagement is eliminated. The elimination is experienced as liberation—and is, in many dimensions, liberation. But the engagement was not merely an obstacle to the commodity. It was the medium through which the commodity acquired its human meaning, and the medium through which the practitioner developed the understanding, skill, and identity that constituted her relationship to the work.
The hearth and the furnace. Borgmann’s paradigmatic example: the wood-burning hearth becoming the central heating system. Both deliver warmth. The hearth demands that someone chop wood, build a fire, tend it through the evening; it centers the household; its warmth is qualitatively different from the even, invisible warmth the furnace distributes through concealed ductwork. The furnace disburdens the user of everything the hearth demanded. The commodity is preserved. The focal experience is dissolved. Applied to AI: Claude Code delivers working code the way the furnace delivers warmth. The hearth model of creative work—the posture of engaging with difficulty rather than delegating it—is not a rejection of the furnace but a deliberate choice to maintain the focal practice alongside the device.
What friction provides. Not all friction is productive. Borgmann distinguishes carefully between obstructive friction—labor without learning, effort without deepened understanding—and productive friction—the specific resistance through which understanding is built, skill is developed, identity is formed, and the centering experience of genuine engagement is found. The challenge the device paradigm poses to every tool that removes friction is the challenge of distinguishing the two kinds and preserving the productive kind while eliminating the obstructive. AI absorbs both kinds simultaneously. The practitioner who understands the distinction must make it herself.
The culmination of the device paradigm. Every previous device operated within a bounded domain. AI is the first device that operates across all domains of creative and intellectual work simultaneously, through a single conversational interface. This universality eliminates the compensatory mechanism that sustained focal practices through previous device-paradigm transitions: the ability to maintain demanding engagement in domains the device had not yet reached. AI has reached every domain. Borgmann called this the culmination of the device paradigm—a difference of kind, not merely of degree, from everything that preceded it.
Focal practices as the response. Borgmann’s answer to the device paradigm is not rejection of technology but deliberate cultivation of focal things and practices alongside it: the activities—cooking, running, making music, the shared meal, the handwritten draft—that resist the logic of commodification because their value is constituted by the engagement itself and cannot be captured by the product they produce. The focal practice survives any device, but only through countercultural effort that runs against the grain of both the technology and the culture of convenience it produces.
The sharpest critique of Borgmann’s framework comes from Peter-Paul Verbeek, who argues in “Devices of Engagement” that the device-versus-focal-thing distinction is too sharp—that some technologies function as “engaging devices,” demanding skill and attention even as they deliver convenience. Applied to AI, Verbeek’s argument suggests that a well-used AI tool demands genuine engagement: the discipline of formulating precise questions, the judgment required to evaluate output, the creative effort of directing the tool toward problems it cannot solve without human guidance. Borgmann’s framework responds: the engagement that Verbeek describes is engagement with the device’s output, not engagement with the material of one’s practice. The developer who evaluates Claude’s code is engaged with Claude’s language, not with the system she is building; the writer who refines Claude’s prose is engaged with Claude’s output, not with the ideas she is trying to think. Both demand attention. Only the latter produces the internal goods—understanding, skill, centering—that constitute focal engagement. A second debate concerns the educational stakes: Borgmann’s analysis of the child and the capacity for engagement is among the most urgent applications of the device paradigm, because the capacities it threatens are the capacities on which every other form of engagement depends. Byung-Chul Han’s smoothness critique and MacIntyre’s virtues framework converge on Borgmann’s diagnosis from different angles.