Soil Well being Now! was held on the Wageningen Worldwide Convention Middle, Wageningen from 8–10 April 2025. It was a 3-day worldwide convention with keynotes, dialogue boards, oral talks, workshops and demonstrations + fieldwork, organized below the auspices of the AI4SoilHealth.eu mission and aiming at bringing the European family of Horizon Europe projects focused on soil health collectively. In comparison with the European Mission Soil Weeks occasions organized by the European Fee, the principle aim of the Soil Well being Now! was to convey collectively researchers and entrepreneurs to the identical occasion and allow them to generate discussions in a extra bottom-up method i.e. let the viewers choose matters and critically and actually deal with their doubts, fears and reservations (as a result of let’s face it: AI or comparable tech won’t mechanically resolve all our issues or present wage for everybody). It ought to be therefore thought of a primarily bottom-up occasion aiming to impress new dialogue and break present dogmas.
We held three knowledgeable dialogue panels specializing in key challenges and alternatives within the discipline of soil well being:
- “The Soil Information Revolution: How Earth Statement, Spectroscopy, and AI are Altering Soil Use Endlessly”
- “Europe’s Innovation Paradox: How Can We Flip Soil Well being Analysis into Thriving Companies?”
- “The Subsequent Era of Soil Tech: Reasonably priced, Accessible, and Farmer-Centered”
To remain as a lot as doable goal & not affect contributors, we now have revealed (unedited) outcomes of polls (Slido) as PDFs through our Telegram channel (https://t.me/opengeohub) as uncooked dialogue notes a couple of minutes after the dialogue panels had been accomplished. Right here you may learn a brief abstract of what has been mentioned and concluded, however we do that simply to make it simpler so that you can discover info — the objects are unsorted and naturally the record could possibly be additional decreased or prolonged. The video-recordings of the periods (nearly 12 hrs of recordings) can be found here. An important takeaways from the convention (in our subjective opinion) are listed under.
AI know-how, particularly automated content material era AI instruments, AI chatbots and automatic navigation techniques (self-driving tractors and comparable) has shortly flooded all areas of life and work. The vast majority of the contributors of the convention agree that the exponential progress of AI know-how options will proceed and at this tempo it is rather nicely possible that it’ll attain 99% of farmers in Europe (about 9M of them) even earlier than the 12 months 2030.
It’s fascinating to notice that there appears to be, nonetheless, two clusters of opinions primarily based on our outcomes: (1) the “optimist cluster” that most likely additionally believes that AI know-how can be fast, profitable and transcend we will think about, and (2) the “cautious cluster” that thinks some greater ranges of experience (knowledge) won’t get replaced quickly sufficient perhaps not even in few many years if ever. Additionally contemplate that information privateness and belief are high of thoughts with regards to AI know-how: viewers voting revealed that information privateness and belief in AI outputs are essentially the most important issues. Panelists emphasised the significance of transparency, management, and “people within the loop” to make sure moral and secure AI use.
Despite the fact that it seems that AI know-how is increasing quickly and may substitute most of the present jobs, ultimately nearly all of present jobs & duties. Does this imply that quickly we won’t have to review and apply statistics (sampling, modeling, statistical testing and many others) in any respect? Is life for the pedometricians and soil scientists turning into simpler so we will all work much less and retire sooner? It may appear counterintuitive, however though AI know-how can velocity up the work it’s going to most certainly require the next information of know-how together with pc science, on high of all primary foundations we get in statistics, soil science and comparable. One of many most popular sayings in 2025 is: “AI know-how won’t take over your job, it’s the folks which are good in AI tech that can take over your job”, so definitively a message for everybody to contemplate that finding out, working towards and mastering advanced scientific fields is right here to remain.
Madlene Nussbaum: “the applying of all these instruments has develop into a lot simpler, so we now have to spend a lot much less time in really implementing machine studying and statistics, however on the similar time we nonetheless have to know what we do and have a transparent image of what the to place the end result into perspective and to really have the ability to have a correct interpretation of the end result and the outcomes, and with out a adequate background information we won’t be able to to place it into perspective”.
One of many fascinating questions raised from the viewers throughout the discussion panel was “what’s the (wholesome) stability between scientific accuracy in information and its utilization or usability?” Greater accuracy will increase manufacturing prices, however many life-saving choices could be made with a lot much less correct / approximate information. There appears to be a normal consensus that the worth of the information is in its use. We must always actually strive our greatest to optimize information accuracy, manufacturing prices and usefulness to match what works for determination makers.
