The hardest part of farming isn’t growing crops or raising livestock. It’s deciding what to do next.
It’s not just because I’ve been watching a Messi masterclass over the past few weeks that I came up with this week’s headline. “Messy job” is a term I heard recently, and it seems to describe farming perfectly.
A messy job isn’t necessarily dirty. It’s a job that is difficult to describe and almost impossible to predict. On any given day, it might involve fifty different tasks, dozens of decisions and a plan that rarely survives until lunch.
AI performs best when a job follows a predictable pattern with clearly defined steps. The easier it is to describe exactly what someone does, the easier it is to teach a machine to do it. Messy jobs are different. They are full of interruptions, exceptions and unexpected trade-offs, making them much harder to automate.
That describes a farmer’s day remarkably well, because no two are ever quite the same.
There is, of course, plenty of literal mess in farming: mud, manure, dust, grease and machinery that normally chooses the least convenient moment to stop working. But the real mess comes from trying to manage biology, weather, people, equipment, markets and money—often at the same time.
Ask someone outside agriculture what a farmer does and the answer sounds quite simple. A farmer grows soybeans, raises cattle or produces milk. That is true in roughly the same way as saying Messi scores goals. It leaves out most of what makes the job hard.
A farmer may begin the morning planning to inspect a field. Before getting there, a water trough has stopped working. He notices one cow in heat and another that looks lame. On a crop farm, rain suddenly appears in the forecast, a grain truck is stuck on the road into the farm and the bank wants one more document before releasing the credit line.
By mid-morning, the original plan has disappeared. Each problem may be relatively simple on its own, but dealing with all of them creates a job that is almost impossible to describe and even harder to predict.
None appeared on yesterday’s to-do list, yet each one changes what happens next. Every farmer reading this has lived some version of that morning. In fact, it is probably what most farmers would describe as a fairly normal day.
Farming is not really one job. It is hundreds of small tasks wrapped around dozens of decisions. Technology has been replacing those tasks for more than a century. Tractors replaced horses. Combines replaced teams of people harvesting crops. Milking machines removed one of the most repetitive jobs on the farm.
Machines took far more jobs away from farming than AI ever will. AI won’t replace the farmer—it will replace parts of the farmer’s job.
As machines took over more of the physical work, the farmer’s role shifted towards managing equipment, people, capital and risk. Farming needed less muscle, but more judgment. And this is exactly where AI can help most.
AI can analyse a crop, predict disease and compare marketing options. But its real value is helping the farmer make sense of all those signals together.
The difficult part of farming has never been finding the answer. It is deciding what to do when weather, biology, machinery, markets and money are all pointing in different directions.
The best agronomic decision may still be too expensive. The ideal planting date may arrive while the planter is being repaired. The most profitable decision may simply involve more risk than the business can absorb.
AI can make those trade-offs clearer, but it cannot remove them. The farmer still has to decide which risk to take and live with the consequences.
Many sectors of the economy are worried that AI will take all the jobs. Agriculture is increasingly asking where the next generation of workers will come from. Across much of the world, there simply aren’t enough people willing to do farm work. Skilled machinery operators are difficult to find, while rural populations are ageing fast.
If AI can remove some of the repetitive, manual and messy parts of farming, the remaining work becomes more valuable. Workers become more productive. Farm managers spend less time gathering information and more time acting on it. Skilled employees become more valuable, not less.
That may also change how the next generation sees farming.
For many, farming has been more of a vocation than a job. Long hours, physical labour and constant firefighting are not especially attractive, even when the family farm comes with a four-wheeler.
But a farming business built around robotics, automation and AI looks very different. Not just because young people naturally like technology, but because the technology allows them to spend more time solving problems than repeating tasks.
Farming will always be a messy job.
The weather will change. Markets will surprise us. Machines will break down at exactly the wrong moment. And biology will continue to ignore our carefully prepared plans.
AI won’t remove that mess. It will simply make farmers better equipped to deal with it.
The future of farming isn’t about replacing farmers. It’s about making farming a job more people want to do.
Thanks for reading.
KFG
Kieran Finbar Gartlan is an Irish native with over 30 years experience living and working in Brazil. He is Managing Partner at The Yield Lab Latam, a leading venture capital firm investing in Agrifood and Climate Tech startups in Latin America.