Send message to Author

Paper short abstract

As GenAI enters agricultural extension, it is framed as a labour-saving substitute. Drawing on workshops with African extension officers, this paper shows how AI redistributes and intensifies advisory work, anticipating the gendered and relational labour that persists in hybrid human–AI systems.

Paper long abstract

As generative AI is introduced into agricultural extension, policy and innovation narratives increasingly position conversational agents as scalable substitutes for human advisors. This paper challenges that imaginary. Drawing on action research workshops with extension officers in Malawi, Kenya, Uganda, Nigeria, and Liberia, I examine how advisory labour is being reconfigured in anticipation of GenAI-enabled systems.

Building on STS scholarship on labour automation, I conceptualise extension as “digiwork”: a blend of symbolic (sensemaking and interpretation), material (technology integration and troubleshooting), and relational (trust-building and negotiation) labour. Through role-play exercises on intent classification, problem diagnosis, and gender-sensitive decision-making, extension officers made visible the tacit, embodied, and context-rich practices required to translate vague farmer queries into actionable guidance. Far from disappearing, this labour intensifies under digitalisation, as advisors absorb new responsibilities for data entry, platform navigation, accountability for algorithmic outputs, and digital upskilling.

The findings show that GenAI does not simply automate advisory expertise. Rather, it risks hollowing out opportunities for practicing judgment while simultaneously amplifying responsibility asymmetries and emotional labour. At the same time, extension officers actively engage in what might be called anticipatory assembling: negotiating which aspects of their work are automatable and which must remain relational and embodied.

By foregrounding agricultural labour within STS debates on automation, this paper argues for hybrid human–AI systems grounded in responsible innovation, participatory governance, and explicit recognition of the social infrastructures that sustain agricultural knowledge.