Dandora

In the AI Story

Dandora receives approximately 2,000 tonnes of waste daily from Nairobi and surrounding areas. The dump operates without formal waste-management infrastructure. Residents of adjacent communities — Korogocho, Mathare, and others — work informally at the dump sorting recyclable materials for sale to middlemen who aggregate them for industrial reuse. The work is dangerous, unprotected, and provides income at or below subsistence levels.

Samasource's recruitment from these communities was deliberate. Janah described the pattern explicitly: workers who had been sorting through waste were, within weeks of training, producing data annotations that trained the machine-learning models powering some of the world's most widely used consumer applications. The contrast was not rhetorical. It was operational evidence for the talent-is-universal claim — evidence that the capability gap the global economy presents as differential endowment is actually differential opportunity.

The specific geography matters for the broader argument. The populations Janah identified as having access to transformative capability once connected to institutional infrastructure were not a small elite of well-educated talent overlooked by chance. They were communities that every global development framework had written off as requiring capacity-building rather than as possessing underutilized capacity. The empirical success of Samasource's recruitment from these communities refuted the deficit framing with data that the framing could not accommodate.

For the AI transition, Dandora functions as both empirical reference and moral provocation. If the talent exists in populations this systematically excluded, and if the tools that could connect that talent to global economic participation now exist at subscription cost, then the question of whether the connection will be made is not a question about the tools or the talent but about whether the institutional infrastructure will be built at the scale required — or whether a new generation of Dandora's equivalents will be told to celebrate their access to tools whose outputs they cannot sell, whose skills they cannot maintain through professional community, whose quality they cannot calibrate against standards they cannot access.

Origin

Dandora was established as Nairobi's main landfill in 1975 and has since become one of the largest waste sites in East Africa. The adjacent communities have grown up as informal settlements that depend on the dump for livelihood.

Samasource began recruiting from these communities shortly after its 2008 founding, with the recruitment strategy reflecting Janah's explicit commitment to reaching the populations most systematically excluded from the formal economy.

Key Ideas

Empirical reference for universal talent. The specific trajectory of workers from waste-sorting to Silicon Valley data annotation provides the most compact available evidence for the talent-is-universal claim.

Deficit framing refuted by data. The success of recruitment from Dandora-adjacent communities refuted the global development industry's treatment of such populations as requiring capacity-building rather than connection.

Geographic specificity. The reference to Dandora rather than abstract 'Nairobi' or 'East Africa' is deliberate; the specific conditions from which workers came are load-bearing for what their trajectories demonstrate.

Moral reference for AI transition. If the tools now reach Dandora's equivalents globally, whether they reach them productively depends on institutional infrastructure of the kind Samasource built — at scales vastly larger than Samasource ever achieved.