James Maisiri argues in Rest of World: I refused to train the AI that could replace me

James Maisiri argues in Rest of World: I refused to train the AI that could replace me

James Maisiri’s essay was published on 3 September 2026. After completing a Ph.D. on technology and agricultural work in South Africa, Maisiri says he was invited to train an AI system to design assessments, teach undergraduates, and mark essays. The job offered income in a setting of high youth unemployment, making the choice a conflict between livelihood and helping build technology that could compete with his work. Maisiri connects his offer to workers elsewhere who train AI or supply data for robots, arguing that both transfer human knowledge and judgment to machines. He warns that African professionals may be recruited to provide expertise at lower cost, while describing how his own teaching experience shaped judgments about what students understand and how their work should be assessed. Maisiri also recounts being interviewed by an AI, then declining to continue after it suggested he retake part of an assessment. He says he still does not fully understand his discomfort or why he walked away. His unresolved concern is what kind of judgment people teach AI systems, including whether they can teach them to be fair, ethical, and responsive to context. He continues to seek academic work teaching students rather than AI.

Publication: 6baceb64-d530-1113-e2b9-ab2b50bc3723

attribution — pass: The reading attributes the first-person essay to James Maisiri in Rest of World's Ideas section, matching the byline and publisher. It is a single argued piece, not a listing. His experiences and arguments are kept as his own assertions.

I refused to train the AI that could replace me By James Maisiri

summary — pass: The summary follows the essay's arc: his Ph.D., the recruiter's offer, the livelihood-versus-principle trade-off, the comparison with other workers, the African cost-incentive warning, the AI-run interview, his unresolved reasons for walking away, and his continued academic job search. It keeps his uncertainty and avoids central jargon. A general reader can follow it. The question reflects his own open questions.

I wish I could tell you my choice was based on principle

date — pass: The dateline reads 3 September 2026, matching originalPublished (2026-09-03T10:00Z UTC). The 2026 unemployment figure predates publication. Nothing is presented as known after publication. modifiedAt (2026-09-01) precedes the publication instant, so it does not indicate a later edit. The page is a publisher page, so no archive disclosure is required.

By James Maisiri 3 September 2026

support — pass: Each claim is directly supported by quoted source text and kept as Maisiri's account. These include the pay and unemployment context, the cost-incentive argument, the AI interview, his not retaking the assessment, his uncertainty, and his ongoing search. The limitations note the lack of independent corroboration. No unsupported benefits, costs or forecasts are added.

For AI companies, there is an incentive to extract expert knowledge from African professionals at relatively lower costs than in Western societies.

media — pass: No media is selected, and no verified candidates were supplied. Leaving media empty is appropriate.