A substantial share of autonomous‑vehicle perception is produced by low‑paid image‑labelers in the Global South who manually tag and verify street images. Their work is routine, precarious, and geographically divorced from the roads the models are deployed on, creating mismatch in risk, pay and oversight.
— Calls attention to data‑labor as an ethical, regulatory and industrial policy issue: AI capability debates must include who does the work, where they live, and how benefits and harms are distributed.
Aeon Video
2026.11.05
100% relevant
Nicolas Gourault’s film Their Eyes interviews annotators in Kenya, the Philippines and Venezuela who parse unfamiliar street images for self‑driving training datasets and describe precarious, low‑paid conditions.
← Back to all ideas