Leveraging LLMs in Agriculture: A Path Forward

In the realm of agriculture, the potential to harness the power of Large Language Models (LLMs) is not just a distant future but a practical reality that can redefine efficiency and productivity today. Taking cues from Shane and Rishi, it's evident that while challenges such as contextual understanding, handling edge cases, and establishing trust persist, the core opportunity lies in empowering agronomists and sales professionals with tools that augment their tasks and action capabilities significantly.

The emergence of LLMs, specifically tailored for sectors like agriculture (eg crop protection and tank mixing) is not far away in my opinion. In combining techniques such as Retrieval Augmented Generation (RAG) and fine-tuning, some pretty incredible results can be achieved. My previous team built a proof of concept in this domain and it showed promising outcomes - this approach is also supported by academic research.

At the heart of this evolution is the belief that LLMs serve to enhance human efficiency (or at least in the form they come today). In my opinion, the narrative is simple yet profound: just as the automobile revolutionized transportation, making horses a less preferred mode, LLMs promise to elevate the productivity of those who embrace them over those who do not. I want to show this transformation with one example of a sales individual managing an extensive territory, where the conventional approach of building farmer relationships through direct visits is becoming increasingly harder due to economic pressures and the expanding scope of territories. So what can we do today?

A Real-World Application: Compassionate, Timely Outreach

Consider the life of a sales specialist/agronomist tasked with fostering meaningful connections across a vast territory. The essence of agriculture sales lies not just in transactions, but in understanding the unique challenges and conditions farmers face - be it weather-related pressures or crop health concerns. Herein lies the opportunity for LLMs: enabling targeted, compassionate outreach that resonates with the current needs of farmers.

Imagine a scenario where understanding the immediate weather implications for each farmer in your territory is crucial. Whether it's anticipating the need for rain, assessing the impact of a recent frost, or evaluating the damage from a hailstorm, staying informed and responsive can be daunting. The data is all available, but how do you leverage it at scale?

This is where the application I developed comes into play. It leverages LLMs and other automation to monitor weather conditions across your territory, alerting you to critical events like frost or significant precipitation at specific locations. More than just notifications, it proposes personalized messages that you can review and approve, enabling you to send automated SMS check-ins or product recommendations that are timely and relevant.

How It Works

The application functions as a bridge between advanced weather monitoring systems and the nuanced understanding of a sales agronomist. By integrating with weather data APIs, the system uses LLMs to interpret weather events and their potential impact on different parts of the territory. When specific criteria are met - say, a temperature drop indicating frost or rainfall exceeding a certain threshold - the system triggers an alert. It then drafts a message tailored to the situation, suggesting a compassionate outreach strategy that acknowledges the current conditions the farmer is facing. This approach not only demonstrates attentiveness but also positions you as a valuable resource, offering solutions precisely when they're needed.

The sales agronomist can then approve or adjust the proposed message and from there it will be automatically sent as an SMS, while also updating the records in your CRM (here Salesforce).

By employing this technology, agronomists and sales professionals can significantly enhance their effectiveness, covering more ground without sacrificing the personal touch that is so crucial in agricultural sales. This is just one example of a potential workflow and underscores the broader potential of LLMs to transform various aspects of agriculture, making tasks more manageable and interactions more meaningful. I truly believe that agriculture is no different than any other industry where LLM’s can significantly impact personal efficiencies, the only question is where do you want it? And the crazy thing is, I didn’t need any coding skills to achieve this nor did it take a long time to build it.

So my question to you is, where do you feel you need to gain efficiencies?

 🚀Big News from Google!🚀

Friends, last week is a monumental day for Google - and for all of us invested in the future of AI. Let's dive into the details.

After much anticipation, Google is showcasing its commitment to the AI race with Gemini Advanced. For too long, I've pondered over Google's hesitance to leverage its unmatched asset: a vast reservoir of user data accumulated through decades via platforms like Docs, Sheets, Gmail, Drive, and Search. The wait, it seems, is over.

What sets a tech giant apart isn't the incremental updates but unique user experience. While ChatGPT dazzles us by curating travel itineraries from simple prompts, Gemini Advanced elevates this to a new level by integrating your personal digital footprint—recent emails, planned destinations, Google Drive contents, and flight options through Gmail/Maps/Drive—to deliver a new user experience.

My Advice:

👉 For those deeply rooted in the Google ecosystem but yet to explore ChatGPT or Claude, Gemini Advanced is a game-changer worth your attention.

👉 For Businesses: While Gemini Advanced is intriguing, it's doesn’t seem ready for enterprise adoption just yet. Continue with Microsoft Copilot or ChatGPT for Enterprise for the time being. I'll update you on any changes.

In reflecting on Google's strategic pivot, it's fascinating to see their integrated approach to AI—a stark contrast to the data silos we've grown accustomed to. This strategy echoes recent developments from Dropbox (Dash and Replay) hinting at a broader industry shift towards more cohesive, user-centric solutions. (Watch this video for more on that!). I really like their vision of the future. 🌐

🕶️ Apple Vision Pro - I wish I would have one!🕶️

It seem as last week many people got their Apple Vision Pro delivered. I’ve watched several YouTube videos and it sure seems Apple knocked it out of the park again. The topic of how we will interact with technology in the future is currently often on my mind. I think I will write about it in another newsletter, but in the meantime just checkout these two reviews 🤯

A bit further out but much sleeker are the glasses below that were unveiled last week.