Close up of Pete, our Peacock
This week I have been collaborating with a team, called Elemental Agronomy, on a venture that is poised to revolutionize large-scale agriculture. The CEO, Bart Peintner, bart@elementalagronomy.com, has agreed to let me tell their story. Full disclosure: I am an advisor and have joined their Angel round.
The problem they are solving is that large-scale agriculture today is dominated by massive equipment and massive overuses of materials such as fertilizer, herbicide, and water. Precision agriculture, as it is called, is not possible with this equipment. It is “brute-force” agriculture.
The Elemental Agronomy team realized that small, inexpensive, autonomous robotic systems could be used to continuously sample the soil and image the plant as it grows. The robot systems could add, on a daily or weekly basis, day and night, just the right fertilizer to each individual plant, at each specific location, as well as remove the weeds, and seed.
To do this requires decisions based on thousands of variables of soil, plant, and environmental data. No current AI system could solve this complex problem effectively.
The team’s breakthrough insight came from the recent groundbreaking advancement in artificial intelligence, known as Large Language Models (LLMs), which can generate human-like language from extensive text data.
They realized that LLM architectures with the thousands of variables of soil, plant, and environmental data can be equally powerful in their use as were the LLMs for text data. And so the team pioneered a deep learning architecture, called Large Agricultural Models (LAgMs) to direct small, inexpensive autonomous robot systems to sample soil, analyze plants, and generate actions such as weeding and fertilizing to optimize plant growth.
The result is a dramatic 30% increase in yield and a 90% reduction in chemical usage.
This has the potential to impact half the world’s food crops, and is a striking application of the new AI capabilities that have recently emerged. Farmers were the initial Angels in the company, with over $500K investment.
The venture revolutionizes large-scale agriculture. I am passionate about it both because it is a new use of the emerging LLM AI capability, and because it represents Precision vs. Brute-force. Frugal vs. Wasteful. Small vs. Large. David vs. Goliath. And the venture has the slingshot.
This is a shorter post than usual, since good weather has come to Northern California. I’ve been busy at my ranch building garden boxes for our vegetables, setting up the watering systems, and getting our chicken coop finished. Our horses, Moon and Ghost, our dog and three cats, and our peacock Pete are enjoying the warm weather.
See you next week.
Your Venture Coach,
Norman
I loved the story showing how tech is not only about the big things, but how important it is to take care of the small details. And I loved seeing to photos... Mike
AgTech is a great field to be in, not just because there are so many ways where technology can make a difference, but by doing so you’re helping the environment and feeding the world. I grew up in the Bavarian countryside, ad even though my industries are different (aerospace/defense) I still have a soft spot for farming. Optimizing resource use through ML and automation sounds like a great idea, and it’s easy to see why you decided to work with them.
Also, love the photos, looks like a truly beautiful place. And I’m still amazed that you have a live peacock!
Thanks for sharing, Norman!