A Scholar's Dilemma: Choosing Humanity Over AI
By Editor • September 3, 2026 • 2 min read
After earning a Ph.D. on the impact of artificial intelligence in South Africa's agricultural sector, I anticipated an academic career shaping future scholars. However, my first job offer was anything but traditional; it came from a recruiter requesting my expertise to train an AI system designed to assess student performance and even teach courses.
Initially intrigued, I found myself grappling with a profound ethical dilemma. The pay of 600 rand ($37) an hour stood in stark contrast to South Africa's minimum wage of 30.23 rand ($2) per hour, in a country facing staggering youth unemployment rates of 47.4% as of mid-2026. This offer posed a challenging choice: earn a decent living by contributing to technology that could eventually replace my role.
This scenario is not unique to me. Many individuals are being recruited to train the very AI systems that threaten their jobs. Highly skilled professionals are often underpaid to enhance AI-generated content, while workers in India are documenting their tasks for a mere 250 rupees ($2.60) an hour, aiding in the development of robots that may ultimately displace them.
Despite the differences in pay scales, our tasks are similar: transferring human knowledge to machines. This trend could significantly impact Africa, where AI adoption is notably low. Most countries on the continent are below the global average of 18% for AI integration among the working-age population, with South Africa slightly better at 23%. Yet, Africa has a youthful demographic and a pool of educated professionals facing high unemployment and stagnant economic growth.
The nature of AI roles is evolving. Unlike data labeling jobs previously outsourced to various countries, the current opportunities require critical judgment and expertise—skills I honed through years of education and teaching. The AI systems would not only replicate my knowledge but also learn my evaluative processes, raising questions about the implications of relinquishing our professional judgment to machines.
In my recruitment experience, I was interviewed solely by an AI, which decided my fate based on a 45-minute interaction. I declined to continue after the AI suggested I retake an assessment, a decision that left me questioning the role of human judgment in hiring.
Ultimately, my choice not to accept the AI training job stems from an internal conflict that I have yet to fully comprehend. As I ponder the implications of teaching ethical judgment to machines, I remain committed to seeking an academic position where I can engage with students, not AI.
Source: restofworld.org
