AI for the informal majority: Designing minimum algorithmic capability infrastructure for Mexico's microenterprise economy

Artificial intelligence policy in middle-income economies often begins at the top: national strategies, cloud infrastructure, startup ecosystems, regulatory alignment, elite talent and corporate adoption. In Mexico, that altitude risks missing the country’s most important developmental unit: the informal and microenterprise economy. For millions of small economic units, the binding constraint is not access to advanced models but the absence of operational memory, financial visibility, managerial routines, bargaining power and trustworthy pathways into formal capability systems.

The monograph develops Minimum Algorithmic Capability Infrastructure (MACI), a sociotechnical design artifact for transforming informal economic activity into cumulative capability. MACI reframes AI adoption as a mission-oriented capability transition: from low-friction capture through voice, images, receipts and messaging, to operational memory, micro-accounting, decision support, gradual formalization, portable reputation and networked cooperation.

Mexico is treated as a hard case because its productive base combines high informality, microenterprise dominance, low managerial formalization, platform dependence, uneven financial inclusion, regional inequality and fragmented institutional capacity. The monograph rejects an implicit AI trickle-down model in which benefits from corporate AI adoption are expected to diffuse automatically toward smaller firms, workers and informal units. It proposes instead an AI for the informal majority: a human-centered algorithmic development pathway in which the success of AI is measured by whether it expands visibility, learning, bargaining power, gradual formalization and local value capture among the economic actors normally excluded from technological upgrading.

The contribution is conceptual, design-oriented and mission-oriented rather than empirically causal. MACI is not a software product, government program or validated impact intervention; it is a transition artifact for expert review, field pilots, comparative testing, institutional coordination and longitudinal validation. Its sharpest design claim is that MACI is defined by where the memory lives, not only by what it captures. As a design criterion, MACI names systems in which operational memory remains under actor control; as a causal hypothesis, otherwise similar instrumentation systems should diverge developmentally according to who controls that memory. Its normative. claim is that Mexico should measure AI success not by adoption volume but by capability built among actors historically excluded from technological sophistication.