Abstract for: Machine in the Loop: Personalizing Farm Advisory for Smallholder Farmers Using AI

Productivity of small farms is crucial for ensuring global food security, yet in many developing countries such as India, significant yield gaps persist across all major crops. Since nutrient deficiencies are a major contributing factor to low yields, bridging information gaps by providing personalized farm advisory can enable smallholder farmers to increase productivity. Our proposed solution is an AI-powered machine that sits in the loop between farmers and traditional extension service providers and leverages modern information & communication technologies (such as mobile phones) to convey personalized information. In this application paper, we present a novel approach that we have adopted in developing an AI-powered information system to deliver personalized farm advisory to smallholder farmers. At each stage of the design process, we conducted field experiments which not only provided us with valuable insights about user needs, but also enhanced our understanding about the causal mechanisms that drive adoption of farm advice by resource-constrained farmers. In the end, such an evolutionary approach helped us design a system that strengthens the abilities of both smallholder farmers and traditional extension service providers in a mutually beneficial way, thereby paving way for a new paradigm of agricultural extension.