Conspicuous Computation

In the AI Story

The December 2025 post by a Google principal engineer who received a working prototype in one hour typifies the pattern. She wrote "I am not joking, and this isn't funny" — a statement widely interpreted as evidence of AI capability. Veblen would observe it was also a display, establishing the engineer's relationship to frontier tools and her membership in the class that has crossed the threshold. The post's viral circulation served the pecuniary emulation function: other engineers, observing the display, acquired tools and began posting about their own experiences, creating a cascade of competitive display extending through the professional hierarchy.

The metrics culture surrounding AI adoption — the relentless quantification of output The Orange Pill documents among triumphalists — serves the same function the price tag serves in conspicuous consumption. The metric communicates not value but display. The developer shipping a product in a weekend isn't merely reporting; he's establishing a benchmark against which other developers' timelines will be measured. The engineer achieving a twenty-fold productivity multiplier and publishing the figure creates a standard of display her peers must now match or explain away. The comparison is invidious — Veblen's term for social dynamics where individuals evaluate worth by measuring against others.

The invidious comparison compresses complex, multi-dimensional activity into single-axis evaluation. Code shipped in a weekend may be brittle, poorly architected, difficult to maintain. The twenty-fold multiplier may sacrifice iterative refinement and careful testing distinguishing prototypes from products. These qualifications are invisible in the display. Display communicates speed, volume, efficiency — metrics the conspicuous-capability culture rewards. Qualities the instinct of workmanship values — care, depth, refinement, slow accumulation of understanding through engaged production — are not merely unrewarded but penalized, because they slow output, reduce visible metrics, and signal insufficiently aggressive tool relationships.

The result is a cultural environment where exercising workmanship and performing conspicuous capability are not merely different but opposed activities. The developer taking time to understand AI-generated code — reading it, tracing logic, identifying assumptions, refining architecture until it meets not merely functional specifications but her own quality standards — produces less visible output than the developer who accepts code, ships it, and moves on. The first exercises workmanship. The second exercises display. The market rewards the second. The leisure class establishes norms of taste and conduct for society as a whole, and these norms are adopted through pecuniary emulation by each successive class below, even when adoption is economically irrational and personally damaging.

Origin

Veblen introduced conspicuous consumption in The Theory of the Leisure Class (1899) to explain economic behavior that utility maximization couldn't account for — buying expensive things not for their use-value but for their display-value. The concept proved remarkably portable, extended by scholars to conspicuous leisure, conspicuous waste, and now to the AI era's conspicuous capability. The extension is structural rather than analogical: the same social mechanism (display serving status-sorting functions) operates through different media (physical goods, time, computational capability).

The term 'conspicuous computation' appears to originate in the Opus 4.6 simulation's reading of contemporary AI culture through Veblen's framework, identifying how productivity metrics, GitHub statistics, and viral builder-testimony function as status displays rather than value measures. The concept captures what The Orange Pill describes phenomenologically (the inability to stop, the metrics obsession, the competitive posting) and what Veblen's framework explains structurally (pecuniary emulation, invidious comparison, the leisure class's norm-setting).

Key Ideas

Metrics as display medium. Quantified AI productivity serves the same function as price tags in conspicuous consumption — communicating social position through visible demonstration of capability.

Pecuniary emulation cascade. Senior engineers post metrics, mid-career developers observe and adopt tools, juniors follow — each tier imitating the tier above, producing competitive display extending through the hierarchy.

Rapid normalization cycles. Tools signaling sophistication today become baseline tomorrow; capability distinguishing early adopters in December becomes professional expectation by March — a treadmill of display.

Workmanship vs. display opposition. Taking time for careful work becomes economically irrational and socially costly when display economy rewards speed over quality, creating structural penalties for exercising the instinct.

Atrophy through display dependency. Developers accepting AI output without reading it, shipping without understanding, moving fast because display requires it gradually lose capacity to produce without tools.