THE QUESTION IS WHEN

By Sebastián Valdés Lutz

Opinion Column

"El Campo" Magazine, "El Mercurio" Newspaper

Posted on April 12, 2021

Agricultural management is measured by its private and social income, based on the maximization and transformation of a series of resources (soil, water, energy, supplies, human resources and mechanics) of agricultural products (fruits and vegetables). The functional transformation is a natural algorithm specific to each farm, which mutates in time, and activates or deactivates variables according to the current context.

The farmer learns from the algorithm that governs his farm based on study, investigation, and most of all from the experience generated throughout the years. The same happens with the advisers, which have spent years and years of searching for new variables, and learning from the behavior of those already known, under diverse agro climatic realities.

Unfortunately, learning based on experience requires the observation of cause and effect to establish the relationship between them, which is an extremely slow process, even more so if it´s not done with the stringency of measurement and control. After years of methodic work, few professionals can, with certain justice of merits, be recognized as experts in specific species or crops. Even so, the algorithm of every farm is so unique and complex, that even the most expert professional will fail in the attempt to accurately foresee with precision its future behavior or unknown scenarios.

The fastest and best way of understanding the algorithm regulating each farm won´t be obtained reading “it´s manual”, or “the case on the Harvard Business Review”, but through maximum observation, registry, study and analysis of its behavior. Not long ago, the marginal cost of controlling a variable was extremely high, given it was normally associated with the physical location of people, manual registries, typing and analysis time, and usually without obtaining a satisfactory sized sample. Today however, precision agriculture provides an extended range of information acquiring methods, which with the evolution of new “vehicles” like drones and satellites, allow the observation and registration in a census form, everything men can, or can´t do, with very low marginal costs. Technology has lowered the cost of learning, in time and money dramatically, and those who have not realized this yet, are losing just that: time and money.

There are many who have heard of what technology changes are provoking in agriculture, and many are also starting to live them. The question is when will technology finally change agriculture as we know it?

Technology is in a state of evolution regarding data gathering and registry of agro climate and agriculture, adding variables and reducing the marginal cost of incorporating them. In parallel, it is also recording the activities that are taking place at the farm, keeping track of equipment and machinery with sensors linking performance and geo-location. Technology is allowing registration of cause and effect, and under predefined parameters, even making decisions about the farm, leaving men to the control and analysis of that which is still beyond its borders. The question is, when will technology reaches those borders?

The information gathered on a farm is still part of each farm’s asset, but when the understanding of the advantages of sharing and being part of a collaborative web that allows the better understanding of the algorithm of each farm, and therefor predicting its future behavior, probably few will see value in keeping this information locked in a vault. No doubt, for this to occur, the global relation of multiple variables won’t fall on men, but on artificial intelligence, which although in its first steps regarding agriculture, is already present. The question is when will agriculture shares the necessary information, so that artificial intelligence may find the relation between all the variables in the algorithms of each farm, and those linking them all?

Once technology has replaced every form of human observation, and controls every registry and information filing method, generates autonomous learning through relating the cause effect variables, then automatically generating farm activities to be executed by robotic equipment in accordance with the events, what we understand as agricultural management will be something completely different to what we know now. The question is, when will we get to that future, if those who govern an agricultural society, have an opinion or vision about it, and what they are doing to face it?

A good subject for the present day.