Inside the AI Factory: the humans that make tech seem human (2023) (nymag.com)

The article explores the critical yet often overlooked role of human annotators like Joe, who work behind the scenes to label and process data essential for training AI systems. With the rapid growth of AI applications, the demand for these workers has surged, as they are tasked with labeling data for everything from self-driving cars to social media algorithms. Despite the repetitive and tedious nature of the work—where tasks can take hours to complete for a meager wage—these jobs provide necessary income in regions with limited employment opportunities. The platform Remotasks, operated by Scale AI, exemplifies this dynamic, connecting annotators with projects while maintaining strict confidentiality about the work and the companies they serve. This hidden labor force is fundamental to the functioning of AI; each labeling effort fuels the models that drive innovations like chatbots and autonomous vehicles. The article emphasizes that the AI sector relies on a complex supply chain of annotation, and as AI systems encounter “edge cases,” continuous human oversight will be essential. This reliance challenges the notion that AI will simply automate away low-skilled jobs, instead suggesting it may alter how work is organized and distributed, leading to new roles that parallel historic industrial transitions. Thus, recognizing and valuing annotation work is crucial for understanding the future landscape of AI and its societal impacts.