PODCAST: “Farmers won’t disappear, but their work will change”
Posted on
“It looks simple, but it’s far more complex than people think.” With that remark, Pieter Blok, Assistant Professor of Data & AI in Agriculture at JADS, cuts straight through a persistent misconception about artificial intelligence in farming. In the latest episode of the Data Dates podcast, he explains how data, computer vision and robotics are reshaping agriculture – and why craftsmanship remains essential.
Growing up with farming
Blok grew up in a farming family in the Dutch province of Zeeland. “From a very young age, you’re immersed in agriculture. That really shaped my upbringing,” he says. That background still strongly influences his work today. “The choices I’ve made in my career are very much connected to that. Yes, my heart is absolutely there.”
After completing his PhD, Blok spent time working in Japan, where he focused on measuring and analysing plant growth using sensors, software and drones. More than anything, that experience sharpened his perspective on agriculture. “Not so much technologically – the differences were smaller than people often think – but in terms of craftsmanship and how deeply embedded it still is.”
From hand-coded rules to data-driven AI
The way AI is applied in agriculture has changed fundamentally in recent years. “Before the AI wave, we did things in a very classical way. As engineers, we programmed ourselves what an algorithm should look for in an image.” Today, that approach has shifted. “Now it’s much more data-driven. You train models on large datasets, and they optimise themselves. You no longer define every internal decision.” That may sound abstract, but the impact is tangible. “We build machines equipped with cameras, software and robotics that take over tasks that are still done manually today.”
Eyes, brains and robot arms
A clear example is selective harvesting. “At the moment, people walk through fields deciding which crops are ready to harvest and which aren’t. That requires a lot of labour.” With AI, that process can be automated. “We use 3D data to determine shape and size. Is the crop big enough? Free of defects?” A robotic arm then takes over, harvesting and placing the produce on a conveyor belt.
Crucially, this is not just about cutting costs. “The main reason is that there simply aren’t enough people willing to do this kind of work anymore. Those hands are no longer available.”
Why scale made sense – and why it no longer does
To understand where agriculture is heading, Blok argues, you need to look back. After the Second World War, scaling up was a deliberate strategy in the Netherlands. “We never wanted to experience hunger again. There was a strong focus on productivity, land consolidation and efficiency.” That approach worked. The Netherlands became a global leader in agri-food. “Many people don’t realise that after the US, the Netherlands is one of the world’s largest exporters of agricultural products.” But the focus on ‘bigger, bigger, bigger’ also created problems. “Monocultures increase disease pressure, chemical use, and contribute to water and nitrogen issues.”
From bigger to smarter
According to Blok, the direction is now changing. “With data and AI, you can move from scale enlargement to scale reduction, without losing productivity.” Instead of vast fields with a single crop, new models become possible. “You can combine crops that benefit each other. Less fertiliser, fewer pesticides, more synergy.” That does require a different mindset. “You adapt mechanisation and automation to nature, rather than forcing nature to adapt to machines.” Smaller, smarter robots play a key role in that transition. “Large machines don’t work well on small, diverse plots. Robotics does.”
The farmer isn’t disappearing
Does this mean farmers are becoming obsolete? Quite the opposite. “The work is changing. I see that with my own parents. My father used to spend much more time among the cows; now he spends more time looking at data.” But the knowledge remains indispensable. “Give a farmer a few data points and they immediately know what to do. That expertise isn’t captured in a model.” Blok compares agriculture to high-tech manufacturing. “In a factory, conditions are controlled. In agriculture, you deal with wind, rain, sunlight, diseases – everything is constantly changing.”
The biggest misconception
If Blok could correct one misconception about AI in agriculture, it would be that it’s easy. “People often think: just add a camera, plug in an algorithm, and you’re done. But every season is different. Every crop variety looks different. Try finding two identical tomatoes in a supermarket.” Automating agriculture in open environments takes years, vast amounts of data and a lot of patience. “That complexity is often underestimated.”
A complementary future
Blok does not see AI as a replacement for humans, but as a complement. “Technology and people should strengthen each other.” Not faster and bigger, but smarter and more sustainable – while preserving the craftsmanship that has always been at the heart of farming. That, he believes, is where the future of agriculture lies.
Listen to Data Dates
Want to hear the full interview with Pieter Blok? Listen to the latest episode of Data Dates on Spotify (in Dutch)
