‘Artificial Intelligence is like Netflix for the farmer’

The Dutch farmer who, for now, isn’t too enthusiastic about Artificial Intelligence (AI) won’t go bankrupt tomorrow if they don’t dive into it just yet. But the smart farmer who does will eventually be able to make money from it, improve their crops, or reduce their environmental impact, predicts ‘AI Professor’ Ioannis Athanasiadis from Wageningen University & Research (WUR) in five questions.

This year, Wageningen University established a new Chair in Artificial Intelligence to explore how AI can be used to tackle complex global challenges. The chair is headed by Greek professor Ioannis Athanasiadis.

How will AI change agriculture in the coming years?

“That depends greatly on where you are. In Africa, for example, farmers use  AI chat bots  groups to get access to basic data (eg local weather forecast), and agronomic advice because they have limited access to information. In the Netherlands, that information is usually available — farmers here already have their basics in order. Here, AI will mainly serve as a decision-support tool. Think of it like Netflix: it recommends what’s interesting for you. If a farmer provides the right data (about soil, weather, yield, diseases, etc.), AI can offer tailor-made advice — just as Netflix suggests films you’ll probably like. It’s all about customization. AI could combine your farm data with that from similar regions, like Belgium or northern France. But it does require you to make your farm data accessible — and, if you wish, share it — to get good advice.”

Will the traditional farmer who ignores AI eventually go bankrupt?

“I don’t think so. But they won’t benefit from it either. If you provide good data, AI can help increase your yield, improve your cultivation, or lower your environmental footprint. Think of Google Maps — it started as just a map, and now it tells you where traffic jams are and how to avoid them. I believe AI will play a major role in risk management in Dutch agriculture. With proper use, farmers can identify and control risks much better. In my team, we develop AI solutions that advise farmers on when and how much to irrigate or spray their fields, helping them optimize production while minimizing environmental impacts such as water waste or chemical overuse.” What’s the first step for farmers who know nothing about AI? “Start by learning a bit about it. There are already plenty of apps that can help. Or you can piggyback — cooperatives are already working with AI, or you can benefit indirectly through machinery manufacturers who are building AI into their equipment. AI can do much more than just translate manuals through ChatGPT. But it’s still hard to predict where things are heading. Try to keep up — learn from each other. You need a critical mass of users to make good analyses; three users of an app won’t get you far. Maybe Dutch farmers have a disadvantage in that sense — margins are tight, and they’re already doing a lot right. But AI might make things just a bit better. It’s like buying a car: there are many options, and you choose what fits you best. The same will apply to agricultural tools and apps — pick what you need and what suits your farm. Artificial intelligence already helps us make better decisions in crop management, precision agriculture, plant breeding, and greenhouse control. On a global scale, AI allows us to make more accurate predictions about the food supply chain and assess how climate change impacts food security.”

Which agricultural sector is AI most suited for?

“The maturity of available AI solutions varies by sector. Generally, AI is more economically viable in production systems with higher added value. We’re working within a large European network, AgrifoodTEF, to test AI solutions in the agri-food sector. Together with over thirty partners across Europe, we test and validate AI and robotics applications in practice — accelerating innovation, supporting SMEs, and giving special attention to startups.”

Can you give examples of AI developments your team is working on?

“We’re developing techniques similar to those used for self-driving cars: AI experiments with different farming practices in simulated environments to discover the most efficient ones. In the Smart Droplets project, we develop smart spraying advice for arable crops to achieve good yields while minimizing the release of chemicals into the environment. We’re also creating a kind of ‘personal assistant’ that learns for each specific field what the best farming practices are — saving resources and optimizing production with very promising results. In the PHENET project, we focus on AI methods for plant breeding — developing deep learning systems that can identify the most efficient genotypes in new environments. This enables us to discover new varieties with higher yields and better quality under changing climate conditions. We’re also collaborating with the Food and Agriculture Organization (FAO) of the United Nations to explore how chatbot-like personal assistant systems can provide recommendations to farmers in sub-Saharan Africa. Building on technologies like ChatGPT and FAO’s extensive library, we aim to develop applications that help train farmers and spread best practices.”