Accepted Paper

Responsible Scaling and Digital Agriculture: Understanding How AI Start-Ups Enact Scaling, Mis-Intelligence, and Dis-Intelligence Through a Multi-Dimensional Lens  
David Kiama (Coffee Research Institute) Ebenezer Ngissah (Wageningen University and Research) Esther Kihoro (ILRI) Cees Leeuwis (Wageningen University Research) Katharine Legun (Wageningen)

Send message to Authors

Paper short abstract

This paper examines how digital agriculture tools must themselves be scaled responsibly to support inclusive agricultural innovation. Using M-Shamba in Kenya, it shows how bundled digital strategies enacted scaling mis/ dis-intelligence over time, shaping success and failure.

Paper long abstract

Digital agriculture initiatives are widely promoted as scalable solutions for addressing persistent challenges facing smallholder farmers, including limited market access, information asymmetries, and poverty. Yet growing evidence suggests that many digital tools struggle to achieve sustained, inclusive scaling. This paper applies a responsible scaling perspective to digital agriculture, arguing that digital tools should be understood not only as vehicles for scaling agricultural practices, but as innovations whose own scaling trajectories require careful, context-sensitive alignment.

Drawing on the science of scaling literature, we conceptualise digital agriculture initiatives as bundled scaling strategies that combine core innovations (e.g. mobile apps, USSD platforms, AI-enabled systems) with complementary innovations such as human intermediaries, organisational arrangements, and communication practices. We introduce the concept of scaling intelligence to capture the capacity of actors to align these innovation bundles with socio-ecological contexts in ways that promote equitable outcomes, while avoiding unintended consequences. Conversely, we identify scaling mis-intelligence and scaling dis-intelligence as processes through which misalignment, exclusion, or loss of relevance emerge.

Empirically, the paper examines M-Shamba, a digital agriculture start-up deployed in Machakos County, Kenya, where it operated for six months through farmer groups and field officers. The case demonstrates how digital agriculture initiatives can simultaneously enact scaling intelligence, mis-intelligence, and dis-intelligence over time. By analysing this temporal unfolding, the paper advances responsible scaling as a novel and productive lens for understanding success, failure, and responsibility in digital agriculture.

Panel P11
Tension? Competing Visions for Digital Agriculture and Rural Development: Smallholder Agency vs profitable business models at scale.