How to use AI for Exponential Landscape Restoration

Exploring the age-old question “What to plant where, and why?

Read also:

  • the case why Investment funds are a key piece of the regenerative agriculture and food revolution (LINK)
  • how funds can do this at (appropriate) scale (LINK)

Could funds also play a role in answering an ancient question which as haunted agriculture for as long as we have been farmers (about 12.000 years)?

What to plant where and why?*

What would become possible if we can start to answer this question not only on a farm scale level (until my farm gate) but at a 
landscape level, ecosystem level, watershed level?

I can hear you think: there are too many variables, local climate, local circumstances, indigenous local knowledge, infrastructure, climate change, climate models for the next 20 years and so many more!

What if a super computer existed to help us answer some of those questions?

With all the developments in deep learning and AI, at some point we will be able to ask the question what to plant and why? So let’s make sure we are ready and go through this thought experiment!

*applies to plants, animals, seaweed etc. What to plant (or walk or float) where and why?

First question

We would ask this super computer when looking at a landscape:

  • Where to start?
    What are the key areas in a landscape from a regeneration perspective? We don’t have unlimited resources, so we should to identify the key pieces of land (nodes) in the landscape/ecosystem first.

Second question

  • What to plant where and why?

How do we:

  • Optimise nutrients per acre
  • Bring down the local peak temperatures
  • Prevent the yearly floods/fires
  • Unlock network effects in restoration
  • Store the most amount of carbon
  • Bring back this specific river
  • Restore biodiversity
  • And last but not least, grow the most delicious and nutrient dens food

We can’t have it all at the same moment (or can we?), so what is the optimal mix for certain outcomes?

These are some of the questions we would ask when somewhere in the (near) future this supercomputer exists.

Until then here are some thoughts I have about what might emerge from the supercomputer answers…..

Exponential growth

I’m really wondering if there is an exponential curve in regenerative agriculture. I know that sounds weird because we are not talking about some software company.

But as the soil is regenerating and the local climate stabilises (less summer heat, more moisture in the air) and many many more positive feedbackloops start kicking in; does biomass on a landscape scale follow an exponential curve before plateauing because of diminishing returns?

Does the speed of regeneration pick up over the years and does produce/livestock and nutrient in the produce follow this trend?

This probably doesn’t mean that the sellable products and financial returns from the land follow the exact same curve, as some biomass will be absorbed by the soil.

For any impact investor interested in regenerative agriculture and food, it is very important to understand if these curves exists in landscapes regeneration and if so, what triggers them!

Farmville

Don’t get me wrong I’m no extreme techno optimist, I’m not saying we should blindly follow some kind of super computers wishes as he (or she) is playing Farmville.

Ancient farmers, indigenous tribes knew in some cases perfectly well what to plant where and why and how to manage large landscapes. But they didn’t have the huge combined pressure of climate changepopulation growth and 2b hectares degraded land (meaning no more other land to move to).

Which means we need to find a way to incorporate and extrapolate that knowledge to a landscape, watershed and ecosystem scale, and unfortunately we need it fast.

We need to acknowledge and work with local native knowledge and augment the new regenerative agriculture practitioners with the latest climate models and prediction software.

Many regenerative farmers are doing this already, but even an enormous farm in Australia isn’t big enough.

Scale

So what does this have to do with investment funds?

Answering all of these questions become easier at scale. A 1 acre regenerative farm influences the hyper local climate a lot but won’t have too much effect a few kilometres away (at least we can’t measure it yet).

I guess also here the complexity of scaling comes in, where the impact of a 1000h regenerative farm isn’t simply 1000x the impact of 1h regenerated land. (Read what I wrote about appropriate scale before)

Which raises another question: is there a network effect in regeneration? As one piece of land regenerates, does that make the neighbouring pieces regenerate faster too? The super computer (in an ideal world) would take this into account and find the pieces of land with the most potential influence. (This will be food for another post in the future:))

Conclusion

As technology keeps moving forward at a breathtaking speed, we might find ourselves in the near future with the possibility to ask some very fundamental questions to a super computer.

My personal question would be, what should we plant where and why? on a landscape, ecosystem level. As this determines if we regenerate 2b hectares over the next decades or not.

If you are building this super computer (or a part of it), prediction software, better climate models, helping farmers to figure out what to plant where and why, please reach out! I would love to learn more!!

P.S.

Some examples which start pointing which got me thinking about this direction:

If you are building something in this space, please reach out! Would love to learn more!!

Thank you Ethan Roland Soloviev for all the feedback!

——————————————

Feedback, comments, suggestions? Reach me via Twitter @KoenvanSeijen, in the comments below or through Get in Touch on this website.

Join the Investing in Regenerative Agriculture and Food newsletter on www.eepurl.com/cxU33P

The above references an opinion and is for information and educational purposes only. It is not intended to be investment advice. Seek a duly licensed professional for investment advice.

Join the Community

2 comments on “How to use AI for Exponential Landscape Restoration

  1. Davide says:

    Very inspiring! Thanks!

Leave a Reply

Your email address will not be published. Required fields are marked *