The Infrastructure of Advantage

In the AI Story

Connectivity is the first prerequisite and the most systematically misunderstood. The discourse conflates availability with adequacy: a mobile phone with intermittent 3G access in rural Bihar is technically a connected device, but it is not a device on which a developer can hold a sustained, iterative conversation with Claude Code. The gap between connectivity that exists and connectivity that enables is vast, distributed along familiar lines — urban over rural, wealthy over poor, Global North over Global South. The tool is available to anyone with internet access; the internet access sufficient to use the tool effectively is available to a substantially smaller population.

Hardware is the second prerequisite. AI-assisted development requires devices with processing power, memory, and screen real estate sufficient for sustained, multi-file, context-rich sessions. A smartphone is not sufficient. The cost of an adequate laptop is modest for a San Francisco engineer and significant for a developer in Dhaka, whose monthly salary may be a fraction of what the device costs. Education — the third prerequisite — requires not merely literacy but the specific capacity to formulate complex problems in natural language with precision and specificity. This descriptive skill is not easier than syntactic skill; it is different, and its distribution is shaped by the same educational and cultural conditions Myrdal spent his career analyzing.

Institutional stamina — the fourth prerequisite — encompasses the supports that enable sustained creative effort: economic stability for time investment without immediate return, legal frameworks protecting intellectual property, financial systems enabling payment collection, and market access for digital products. Each of these institutional prerequisites is weaker in the periphery than at the center, and the weakness of each reinforces the weakness of the others.

The closest historical parallel is the Green Revolution. High-yield crop varieties developed in the 1960s and 1970s were technically transferable and promised agricultural transformation. The benefits accrued overwhelmingly to farmers who possessed institutional infrastructure — credit for inputs, secure land tenure, irrigation, market access. Farmers who lacked these conditions were left worse off, as increased production by their better-resourced neighbors drove down crop prices while input costs rose. The technology was democratic in availability and anti-democratic in impact, because impact was mediated by institutional conditions the technology could not create. The parallel to AI is uncomfortably precise.

Origin

The specific phrase "infrastructure of advantage" emerges from the Myrdalian reading of Segal's imagination-to-artifact ratio, but the underlying analytical move — examining the institutional prerequisites that determine whether formal access converts into substantive capability — runs through Myrdal's entire corpus, from An American Dilemma (1944) through Asian Drama (1968). The framework anticipates Amartya Sen's conversion factors by three decades and operates with comparable analytical precision.

Key Ideas

A chain is only as strong as its weakest link. Tool-access is one link among many; the chain fails at the weakest prerequisite.

Availability is not adequacy. Technical connectivity differs from connectivity sufficient to enable the use it theoretically permits.

Descriptive skill as new gatekeeping. Natural language interfaces do not eliminate skill requirements — they change their character.

Institutional stamina as hidden precondition. Economic stability, legal protections, and market access determine whether tool-access becomes sustained capability.

Green Revolution precedent. Transferable technology in non-transferable institutional environments concentrates benefits among the already-advantaged.