REAP Conference 2024 registration is open
Book tickets, feature in the technology exhibition or apply for a REAP bursary - available for farmers and those in full-time agriculture-related study

From algorithms to AI – ChatGPT and agri-tech

Agri-TechE Blog
Agri-TechE

With high profile public commentary about the future of artificial intelligence (AI) in society, this month we’re chatting about ChatGPT and its role in agri-tech. 

In case you haven’t tried it, ChatGPT is an advanced AI language model that can generate human-like text and engage in natural language conversations (so it told us).

AI is no stranger to agriculture; increasingly technologies exploit algorithms that have “learned” on real-world data how to make decisions. From early estimates of harvest yield to discerning weed from crop for precision applications, AI is already making ways into everyday agriculture operations.

But ChatGPT is a potential game-changer as the most well-known of the emerging “generative AI”.

Some versions are (currently, at least) free to use, widely available and learning rapidly. You simply type in a challenge, question or task and it types out the answer in front of your eyes. The exciting thing about ChatGPT is that its dataset is the whole of the internet, in effect – some would argue – all human knowledge.

How can ChatGPT’s AI benefit us?

Technology advancements like this make knowledge more accessible to everyone. New users can interact with these algorithms without training. You no longer need to know coding to be able to get a bespoke data analysis based on your criteria.

And the answers can tap into new, creative avenues. Ask ChatGPT for the plot holes in your favourite film, and I’d be surprised if it doesn’t find something you’ve never thought of, no matter how many times you’ve seen it.

What if we ask it what we haven’t studied in soils? Or if there’s a pattern in the data between two of our sensors? Do solar flares affect our milk production? Chat to it further, and it can give ideas on how to investigate the answers.

Then there’s the enormous time-saving potential. Could an agronomist create a bespoke dashboard for a client’s farm-specific needs in a matter of minutes, just by typing a few sentences?

Sounds too good to be true…?

As with all innovations, there are inherent risks.

For example, it can be pretty difficult to find the source for answers provided by ChatGPT. How do we know this information is reliable enough? Or current? And not ‘learnt’ from some unusual or out-dated options contained in the web with little scientific support but presented by ChatGPT as a fact like any other.

Plus, when technology is so fast moving, regulating it is extremely difficult. To some extent, we are relying on the developers to test it, self-regulate, and ensure ethical development. Frequent scandals of badly trained algorithms – such as the GCSE results tool used by the UK government, or the consistent bias against ethnic minorities in crime-reduction – highlight the need for constant real-life assessment and testing by experts in the field.

Although users need no training to ask questions, knowing what prompts to use and how best to use them is key. ChatGPT is just a tool like any other – if we don’t know how to use them effectively, they can’t give us useful outputs!

Will ChatGPT replace us?

Given time, ChatGPT and its successors will probably design the world. What then of the role of the expert vet, the agronomist – even the UK’s longest established agri-tech network organisation?

No-one has a monopoly on information – almost everything is “out there” if you know where to look, which YouTube videos to watch, what website to search and who to talk to. But who has the time for all this research and information gathering?

This is where ChatGPT and its friends will no doubt be part of life going forward – to aid and augment.

For Agri-TechE (and our members), we see this is as a tool to help us scale our activities, be even more effective, and provide additional insights – but not as a replacement.

ChatGPT will give a comprehensive answer to the question you ask; Agri-TechE will respond to your question, set up a bespoke introduction, and give you useful and relevant answers to the questions you didn’t even know to ask yet… So says Andrew Francis , Senior Farm Manager at Elveden Farms:

I take great value from events like REAP because often you get ideas and come away with something that you didn’t go intending to find. In some ways you’re being presented with the answers before you have considered the question.

The REAP conference was fantastic and I would liken it to having a room full of answers floating above you and just needing to grab the right answer to match your question – sometimes before you even know the question!

After the flood making information beautiful

Meet the Network
Agri-TechE

Nick Cross, After the floodTo survive in a complex world humans have evolved the skill to extract information quickly from patterns. This ability is exploited by infographics, graphics that display data as pictures. Familiar examples include the London Underground map, which supports navigation, and the Met Office maps, which allow weather prediction.
After the flood is now taking this type of data visualisation to a new dimension with artificial intelligence. It takes insights from multiple interactions (people-machine and machine-machine) and displays the findings as deceptively simple dashboards.

Personalised food production 

After the flood’s Chairman Nick Cross (pictured right), who also manages his family farming business in Suffolk, explains: “Traditional data analytics are based on collecting data and then providing retrospective insights.
“We are moving into a new era of active data that uses real-time data to provide intelligent services. Perhaps there will come a time when food production will be personalised!”
Instead of creating static images using historical data, After the flood creates a dynamic interpretation of live data. This allows fast reactions and the ability to create systems that learn from experience to respond to changes in their environment.
Within agri-food this could be using customer buying behaviours to predict demand for perishable goods, or monitoring fungal spores and weather conditions to allow preventative, precision spraying.
Nick continues: “I think there will be exciting opportunities to create intelligent data flows between customers, stores and the producers themselves, allowing farmers to be more responsive to specific consumers’ tastes and dietary needs.”

Information is beautiful 

After the flood has won an ‘Information is Beautiful’ award for its London Squared Map (see below)
London squared - After the flood
In partnership with Future Cities Catapult, After the flood turned London’s boroughs into a choropleth – or series of shaded cells – that can contain numerous types of data more effectively than a traditional geographical map.
The London Squared Map allows data to be compared across boroughs very easily, and governmental and other agencies are able to edit it to create their own versions for their web pages; for example the London Fire Brigade is one of several organisations that now use it. Some versions even place Instagram images inside the squares.

Managing for the public good

Nick says: “Who knows how the application of AI will really play out with farmers at the sharp end. If you go back 15 years or so, few people would have predicted the impact of cameras on mobile phones and how this has changed the way we live and interact.
“There is already a lot of noise about the role of AI in developing efficient crop production techniques.
“My guess is that there will also be exciting opportunities as farmers take on the challenge of managing land resources for the public good – because we will have to develop a much more sensitive, localised and dynamic picture of the impact we’re having on things like soil, climate change and invertebrates.”

Find out more 

Nick Cross will be presenting at the Agri-Tech East Pollinator “The AI’m of machine learning in agriculture” on 20 February 2018 at The Sainsbury Laboratory, 47 Bateman Street, Cambridge, CB2 1LR.