You On AI celebrates a specific figure: the AI-augmented solo builder who ships what once required teams, the entrepreneur whose imagination-to-artifact ratio has collapsed to the width of a conversation. Alex Finn — 2,639 hours in a year, building a revenue-generating product without a team — is the paradigmatic example. The narrative frames this as individual empowerment: barriers dissolved, capability distributed, the bottleneck broken. The claim is real but partial. Finn's building depended on Claude, on the training corpus that shapes Claude's outputs, on the API infrastructure that made interaction possible, on the cloud systems that hosted the product, on the payment processors that handled revenue. The collective that made the solo work possible was vaster than any team Finn could have assembled in pre-AI conditions. It was also less visible.
The visibility asymmetry matters. In the old network, the collective was composed of team members with names, faces, desks, roles. The designer sat next to the developer. The project manager coordinated. Contributions were individually recognizable, individually credited, individually compensated. In the new network, the collective has been compressed behind an interface. Its members — the training data contributors, the infrastructure operators, the semiconductor workers — are not visible to the user. Their contributions are channeled through systems that efface their individual identity, and the economic and social structures that recognize contributions do not extend to them.
This has immediate consequences for credit, compensation, and recognition. When the solo builder sells a product, revenue accrues to the builder. The collective receives compensation through different channels: API fees, subscription payments, wage labor. But the connection between the collective's contribution and the specific artifact is severed by the black box. A developer whose open-source code trained Claude and whose work therefore contributes to every Claude-assisted artifact receives no royalty on those artifacts. The cobalt miner whose labor made the GPU possible has no standing in the economic network that the GPU enables. The economic structure has not caught up with the collective structure.
Dylan's 'Like a Rolling Stone,' which the You On AI uses as an extended meditation on creativity, illustrates the pattern from the other direction. Dylan was not the source of the river but a stretch of rapids through which cultural tributaries converged. The song was synthesis — Guthrie, Johnson, the Delta blues, the Beat poets — flowing through a specific biographical architecture. The individual contribution was real and specific. But the contribution was constituted by the network, not independent of it. Extend this to the AI-augmented builder: her contribution is real and specific, but it is also constituted by the invisible collective that her tools make possible. The question is not whether the collective exists but whether it will be acknowledged.
The idea runs through Latour's entire career, with roots in his early laboratory ethnography. In studying scientific discovery, he repeatedly observed that what counted as 'individual' achievement was the visible residue of collective work — technicians, instruments, funding structures, prior research — that the rhetoric of scientific genius systematically effaced. The scientific paper, with its single or small-group authorship, is a genre that performs the effacement.
The concept has been developed further by scholars working on what they call 'ghost work' — the human labor (content moderation, data labeling, training feedback) that makes AI systems function and that is systematically invisibilized in the presentation of the systems to end users. Mary Gray and Siddharth Suri's Ghost Work (2019) traces the specific labor conditions of the invisible collective that sustains large-scale AI deployment.
Solo builder as visible node. The individual celebrated in the AI empowerment narrative is the most visible participant in a much larger network, not the sole source of the output.
Collectives have not shrunk; they have been concealed. AI-augmented work depends on collectives vaster than any pre-AI team, but the collectives operate behind black-boxed interfaces.
Credit asymmetry. Visible nodes receive credit; invisible contributors receive compensation through different channels, if any. The economic structure does not reflect the collective structure.
Ghost work. Content moderators, data labelers, and training-feedback workers — often in low-wage contexts in the Global South — perform the labor that makes AI systems functional. Their invisibility is a structural feature, not an accident.
Recognition as governance. Making the collective visible is not merely accurate description. It is the prerequisite for governance structures that can address the collective's contributions, compensations, and working conditions.