Upstream Ag Professional - October 15th 2023
Essential news and analysis for agribusiness leaders
Welcome to the 13th Edition of Upstream Ag Professional!
As Upstream Ag Professional members, you receive access to exclusive videos, including monthly reviews of the most important news, events, research, and deep dives from around the agribusiness and agtech landscape. For September’s summary, the topics include:
AGCO Acquisition of Trimble Assets and Joint Venture
Indigo Ag’s $250 million raise
Psychology and Incentives: In Inputs and AgTech Adoption
Rantizo’s $6 Million Raise and Ambitions for Automation Orchestration
You can check out the full video here:
Upstream Ag Professional September Summary Video
Index for the week:
The Shortcomings of Large Language Models for Agronomy
The Challenges of Nitrogen Fixing Microbes and What it Means for R&D and Commercial Efforts in the Industry
Notable Biological Commentary Round-Up
Bioherbicide Gap and Potential in Market
AMVAC’s Bob Trogele Interviews Pacific Agriscience’s CS Liew About Biorational Company Successes and Advice for CEOs
FMC Reveals Insights on Biologicals Market Growth in Latin America
The Big Seven of the CropLife 100: The Evolution of America’s Largest Ag Retailers
Nitricity Commissions New Pilot System
Farmers Business Network Accelerates Work with ADM to Expand Regenerative Agriculture
Creative Precision Spraying Image
1. The Shortcomings and Opportunities of Large Language Models in Agronomy - Upstream Ag Professional
Key Takeaways
Large Language Models are improving to the point that they can attain upwards of 93% on agriculture industry exams, including the CCA.
These improvements illustrate the capability of LLM’s as assistants to agronomists moving forward. The efficiency of agronomists will be highly enabled through LLM’s in certain realms, including basics of crop protection information.
These have shortcomings when it comes to replacing agronomists because of the realities of human trust, a lack of contextual awareness and the importance of edge cases in agronomic value creation.
This week, I read the following research paper on using LLMs to pass agriculture-based exams and thought it was well done:
GPT-4 as an Agronomist Assistant? Answering Agriculture Exams Using Large Language Models - Arxiv
To demonstrate the capabilities of LLMs, we selected agriculture exams and benchmark datasets from three of the largest agriculture producer countries: Brazil, India, and the USA. Our analysis highlights GPT-4’s ability to achieve a passing score on exams to earn credits for renewing agronomist certifications, answering 93% of the questions correctly and outperforming earlier general-purpose models (GPT-3.5), which achieved 88% accuracy. On one of our evaluation datasets that had published student scores, GPT-4 obtained the highest performance when compared to human subjects.
Given the increasing capability, I think it is worth exploring the implications for agronomists and farmers.
In May, my friend Rhishi Pethe had similar curiosities to the researchers as to whether a Large Language Model could pass the certified crop advisor exam.
It also grew my interest as to whether LLMs could pass the CCA exam. I passed the CCA exams in 2015 and dug into some old practice tests to see how ChatGPT would do.
In a much less scientific matter and with a smaller data set than the above study, I asked ChatGPT (on model GPT-3.5) ten questions— it got seven right or 70%, which would typically be enough to obtain CCA accreditation in any given year— similar, but slightly lower than most of the examples in the paper.
The paper states that the “study offers a unique perspective and contributes to the understanding of AI’s potential impact on agriculture by providing a baseline for future benchmarks about the use of large language models to solve agricultural problems.” The researchers do not attempt to call out a replacement for agronomists, and I think it’s worth highlighting why that’s important.
Any CCA will tell you the exam establishes that you have a baseline understanding of crop production. There is, however, extreme nuance when making objective recommendations to farmers.
Therefore, I asked ChatGPT follow-up questions to dig deeper into some of the questions it got right.
In this example, the number it suggests is relatively low compared to generally accepted numbers, and I found the same with other follow-up questions. The answer fell short of the needs a farmer would have (though it does deliver some additional interesting insight). This illustrates that when it comes to making specific recommendations, it still has room for improvement.
Knowledge is helpful, but being able to apply knowledge in a practical way is where agronomists create value. I find it interesting to think through why it would be challenging for an LLM to move beyond an assistant or “co-pilot” for the agronomist.
There are many reasons there will be challenges for an LLM to ever move beyond an assistant for an agronomist, but three stand out:
Human Trust
Edge Cases
Contextual Awareness
In the full article accessible above, I break down the realities of these three reasons limiting LLMs as a replacement, and what the implications are for use cases of LLM’s in agronomy, along with what it means for changing up the role of agronomists in the future.
For more on LLMs, their shortcomings, and where there might be a fit, check out these previous installments within Upstream Ag Professional:
Artificial Intelligence and the Supply Chain in a World of Converging Agribusiness Software - Upstream Ag Professional
ChatGPT Implications for Agriculture - Upstream Ag Insights
2. The Challenges of Nitrogen Fixing Microbes and What it Means for R&D and Commercial Efforts in the Industry - Upstream Ag Professional
Key Takeaways
Crop-agnostic, nitrogen-fixing microbes present a significant opportunity within agriculture for investors, agribusinesses and farmers.
There has been shortcomings when it comes to consistency of performance when looking at third-party data along with anecdotal farmer conversations. This doesn’t mean the products do not work— it illustrates that there are challenges that need to be overcome, including surrounding microbe fitness and N transfer from the microbe to the plant, which was highlighted in a recent microbiology journal paper.
These challenges present a view of where R&D resources are going and are likely to go from large incumbent entities that want to unlock the value of crop-agnostic nitrogen fixation— including specific gene editing in crops as well as in microbes.
This week I read a recent journal article from Trends in Microbiology touching on the dynamics of nitrogen-fixing organisms in agriculture, pointing out the significant scientific challenges that companies are working to overcome to deliver commercially viable crop-agnostic, N-fixing products into the market.
I think there are points in this article that contextualize why many products see inconsistent responses in a field setting.
Crop-agnostic, nitrogen-fixing biological products have huge potential in agriculture.
Crops like corn or wheat fixing even a portion of their Nitrogen needs can increase yields and quality, decrease deleterious environmental impacts of synthetic nitrogen fertilizer plus streamline logistical costs and efficiency across a farm operation by minimizing application needs.
The agribusiness and venture capital world knows the potential.
The maker of nitrogen-fixing product ProveN 40, Pivot Bio, has raised over $617 million to date and is on millions of acres.
Sound Agriculture, the maker of Source, has raised over $170 million.
Kula Bio (an organization that is pre-commercial) has raised over $74 million.
Corteva Agriscience purchased biological company Symborg in 2022 ( for an undisclosed amount), a company whose most prominent product has the active ingredient Methylobacterium symbioticum, a nitrogen-fixing bacteria.
Developing a crop-agnostic, nitrogen-fixing product is incredibly challenging, though. As early as 1917, scientists attempted to cultivate the rhizobia from legumes and inoculate these into other crop species. To date, however, none of these attempts to transfer the complex root nodule to non-legume plants has succeeded.
This is why much of the effort has been focused on free-living diazotrophic (nitrogen-fixing) organisms.
There are many significant challenges to overcome before we see a point where crop-agnostic, Nitrogen-fixing microbes become a standard application. Even though strides have been made with several commercial products above, they are still imperfect at delivering a response consistent with expectations and economic demands, as an NDSU study highlighted earlier this year (I dove into the dynamics of it here: Upstream Ag Insights - April 30th 2023).
The aforementioned study highlighted some of the specific challenges surrounding diazotrophs I called out in April (image in article) that are worth digging deeper into.
In the article I dive into two key problems:
Getting nitrogen fixed by the microbe into the plant efficiently.
Microbe “fitness” and their ability to compete in a hostile environment.
There are opportunities to improve in both of these areas, and companies working on it today.
In the full Upstream Ag Professional article linked above, I go into both of the above aspects in-depth including the science behind and what the implications are for company R&D and commercialization, along with highlighting what a start-up is doing to navigate the realities of fitness penalties in nitrogen-fixing microbes.
3. Notable Biological Commentary Round-Up
Key Takeaways
Bioherbicide cost of goods is trending down, meaning more opportunity with bioherbicides in the future.
There is nowhere in agriculture where consolidation and attrition is likely than in the biological space.
Over 15% of the biologicals used in Brazil are manufactured on-farm.