Creatures and Machines

urban sustainability, resiliency + humans


ICT + Ag extension

Last week I attended a conference on information and communications technology (ICT) and the opportunity it presents to improve smallholder agriculture in the development context. Gathered vendors offered a variety of ideas surrounding mobile credit service access and educational ag extension programs, intended to increase overall crop yields, market productivity and profitability. By far the most interesting discussion centered around an afternoon panel on the future of ag extension and the value of human interaction and the element of trust in decision models.

Agricultural extension, or ag extension, consists of a chain of workers sharing and disseminating new scientific research and best practices on crop management, in an effort to bring about the best yields relative to annual forecasts. Ag extension workers regularly help smallholder farmers make important decisions on fertilizer application, seed selection, soil management, timing of planting, as well as post-harvest production issues, such as grain storage and price procurement. Advances in ICT connectivity, and access to data and stronger climate models, greatly expands the amount of knowledge ag extension workers and farmers have available through their mobile phones, allowing them to potentially better manage their network of farms. With the abundance of data and information available, it is important to understand the value that humans add in distilling relevant data sources and applications to provide sound agronomic advice.

Kentaro Toyama, author of Geek Heresy: Rescuing Social Change from the Cult of Technology, highlighted the role of trust in this decision process, especially with the explosion of technological solutions on today’s market. He offered the example of both travel and insurance agents, and how their role had greatly diminished over the last decade as people have relied more heavily on data aggregating websites to make cost-effective purchasing decisions. Does the incremental value add of trust through human interaction reduce to marginal significance in smallholder agricultural development? A participant reminded us that a decade ago, most consumers were hesitant to store credit card information online.

Today, a large percentage of the population uses internet payments and savings mechanisms in good confidence that their money is well safeguarded. In Africa, and other developing regions, mobile money is one of the most important forms of currency.

Agriculture has been around for millennia, but still the notion of “precision agriculture” remains fraught with risks and uncertainties for a variety of reasons. While applications and big data can offer both farmers and ag extension workers patterns and algorithms to better contextualize past issues and inform decision making – humans remain an irreplaceable factor in the overall decision making process because of the importance of non-data elements in agricultural performance – the same way an entrusted doctor can offer more value than WebMD.  As time continues to build our confidence and trust in data driven decision modeling, when evaluating forecasting and crop growth, the best applications on improving ag extension services will continue to account for the human factor rather than try to prove it obsolete.

While it remains unknown what eventually causes developing, agrarian societies to make the leap to developed, post-industrial economies; it would be myopic to believe that ICT applications can single-handedly solve this problem through ag extension services. In an interesting conversation between Tyler Cowen and Jeff Sachs, where Professor Sachs highlights the importance of geographical contextualization in agricultural development, stating, “you cannot simply take, most of the time, something that worked here and plant it here”. Sachs points to the double miracle of the Green Revolution, where Mexican seeds worked in Indian growing conditions to significantly improve crop yields, but this success was largely contingent upon a year of experimentation by Norman Borlaug.