Exploring Digital Farming: An Interview with Frido Hamoen from Lely

Following the webinar we recently held, we conducted an interview with Frido Hamoen from Lely. Below, we delve deeper into the topic of Digital Farming and Frido's vision on it.

How can I prepare myself as a farmer to be able to implement AI tools asap when they are introduced? 

The development around sensors, data and AI algorithms are emerging. This offers both opportunities for farmers, but there are also risks. This is no different than with any other new development or technology. My advice would be to get yourself informed about the opportunities and risks of AI. This can be done by reading about AI in magazines, newspapers and on the internet, talking about AI with your advisors and other farmers. But you can also talk to your children or young people working on your farms as they are in general very quick at adopting new technologies and learn quickly what the pros and cons are. 

How can I trust the outcomes of AI models? 

This really matters in the way the AI models are trained. What data is used to develop the AI model? And what data is used as feedback to constantly train the model further. It is all about the quality of the data. How reliable is that data? In what circumstances is this data collected? Are these circumstances similar to yours? How big is the dataset and how fast is new data added to that dataset? Is the model supervised by experts that check if the outcomes are making sense? Who developed the AI model? How credible is that organisation? How was the AI model validated? How many farmers are already working with the AI model? A lot of questions and for many farmers too much to handle. So also here, please talk to your advisors and other farmers about this. And experiment with the AI models to teach yourself what they can and cannot offer for you. Many models are offered with a free trail period or can be used for free with some restrictions. By doing these experiments you can assess the value these models bring to your farm. 

What is the difference between strong AI and weak AI? 

We talk about strong AI when the model can learn something in one context and apply these learnings in a different context. This comes close to how the human brains work. That said, strong AI systems don’t exist yet. 

What can be the value of copilots for farmers? 

A good working copilot (which is well trained with high quality data, extensively validated and supervised by experts) can add a lot of value for farmers. You can save time, increase your production and decrease your costs by using them as the copilot will make sure that you take better decisions in less time. So, the potential of AI technology is huge! 

What skills does AI require for farmers? 

Main skill is that you need to be open for change. Some outcomes of AI will be surprising, and it will not always be clear why AI did get to that outcome. This will raise some eyebrows when you start working with AI. Make sure that you keep the overview and remain in control of what is happening at your farm by monitoring the results. And finally, you need to be able to accept that AI will not be perfect from the beginning. But are all decisions taken today by humans perfect?