Generative AI in Agricultural Management

Peter Gill

Managing Director

The Future = Generative AI

Since 2017 the global management consultancy McKinsey has been conducting a global survey into the use of AI by businesses. The 2024 findings are the most significant because they reveal the differences between 2023 and 2024 - the first two years in which Generative AI was being used (November 2022 being the launch of ChatGPT). Previous iterations of the survey were largely restricted to focussing on the use of machine learning AI (ML).

The above comparison of the 2023 and the 2024 results show 44% of businesses using Generative AI reporting an increase in business revenue and 39% reported a reduction in costs. The mean averages amongst those companies were respectively revenue increases of 5% and cost reductions of 10%. It is important to note that this is still the very early days of the application of Generative AI.

The technologies are advancing at pace. Incredibly powerful features have been unveiled since November 2022 - many of which were wholly unachievable as recently as ten years ago. Big new players, not wanting to miss out in the AI goldrush have joined the race.

Generative AI and Business decision-making

You may be familiar with a notable study into farm profitability conducted by the AHDB with Andersons, the farm business consultants, and published in 2018.

The notable thing about the study was that it compared matched types of farms - with the difference being their profitability - to try and find reasons how some were succeeding in making profits and others not.

Thinking more widely, the same kind of study would be good if applied in other sectors of business: property developers, professional services agencies, garden centres, marketing agencies, hotels and so on - the reason it was possible for farming was largely because of the existence of ongoing govermental detailed farm business surveys.

A headline finding of the AHDB/Andersons study was remarkable: top 25% farms were making £100,000 more per year than those in the - matched farms, remember - bottom 50%.

The report also cites a figure of only 5% of factors affecting farm performance being out of the farmer's control. It would be contrary to imagine that business profitability in many other businesses outside agriculture is not likewise almost entirely down to management's business decision-making.

Stepping Up To Using Generative AI

The idea that Generative AI could autonomously largely manage a business or a farm isn't as crazy as it sounds (although getting the senior management on board is a readily foreseeable major challenge).

What is already eminently do-able and almost certainly best business practice for every business is to start using Generative AI as management co-intelligence. The great likelihood is that the resulting 'joint decisions' would steer profitability upwards, particularly so for the least profitable businesses. At the least an input of fresh management thinking would be highly beneficial.

In five years AI will have become a part of the critical infrastructure of your business.

Is this realistic? Let’s assemble some facts. Actually, we just need one fact: AI performs better than humans in every area of human expert knowledge studied.

An example involving absolutely mission-critical AI success showed a cutting-edge medical diagnosis system, costing millions to develop, bested by a fresh-out-of-the-box plain vanilla AI. This is the new norm.

One of several physicians v. AI studies showed:

AI flying solo - 90% accuracy
Expert using AI - 76%
Expert alone - 74%

For business users of AI and managers a profound observation was made. The doctors using AI performed less well than AI alone because they tended not to heed its conclusions when they didn’t agree with them. That is interesting.

Most people testing out AI have looked over the free version of ChatGPT. In point of fact there are over 20 rapidly evolving state-of-the-art (SOTA) systems—the outcomes of over a trillion dollars of investment, and counting. Some AI systems are free (for now at least), others are currently $200/user/month.

AI Team-Training

Knowing which systems are best for your business’s needs, optimising working across multiple AIs and how to verify outputs are core components of our AI team-training programme. Further modules include data security, intra-organisation collaboration, including the building of time-saving AI ‘in-house apps’, and integration with your office and business workflows.

Inside and outside of business and across the world the use of AI is undergoing exponential growth so holding up on business adoption is wiring in a future need to make up ground lost to agile competitors.

To make the most of your team’s time our AI training will be bespoke to the functions and processes of your organisation—just get in touch with us now for more on how we can be working together soon.