In the village of Ndodo, 40 kilometers south of Malawi’s capital, Lilongwe, farmers gather in the shade of an acacia tree as a voice through a smartphone tells them how to get rid of a weevil that is destroying their sweet potato crops.
The advice offered by the app in the local Chichewa language is one of the first examples of how artificial intelligence is being used to help subsistence farmers in some of the poorest parts of the world.
Artificial intelligence is increasingly used in agriculture in large areas of the developed world.
Piloted by Chicago-based non-profit Opportunity International, the app called Ukanzigi – which translates as “Advice” – runs on WhatsApp and uses data from ChatGPT and the Malawian government’s English-language agricultural handbook to answer questions or to diagnose crop and livestock diseases.
Most of the people in the area cannot read or write. Government farmer support agent teaches implementation to farmers. In this way he helps the locals write down problems they have on their farms and reads them the answer they receive. They also have the opportunity to hear the answer through voice notes that come in their own local language.
The app could become critical for the 3.1 million smallholder families who rely on agriculture to make a living in this landlocked country in southern Africa. Its use comes after the devastating Cyclone Freddy hit Malawi in early 2023, displacing nearly 100,000 people, leaving fields flooded and nearly wiping out the annual soybean crop.
Although artificial intelligence is increasingly used in agriculture in large parts of the developed world, from China to the US and Europe, its emergence in poorer countries and for subsistence farmers is relatively new. The success of the Ukanzigi application in Malawi – where small-scale agriculture provides livelihoods for more than 80% of the country’s 21 million people – may pave the way for its introduction in other parts of the world. There are 600 million smallholder farmers worldwide, who grow a third of the world’s food.
Cyclone Freddy showed how quickly a climate disaster can destroy an economy and food system, and farmer support agents said Ukanzigi could have helped them not only cope with Freddy’s aftermath but also prepare for it.
The pilot program has just finished, and Opportunity is collecting qualitative data through surveys and focus groups to fine-tune the app as it looks to expand its reach.
Given the fact that the app has networks of farmer support agents across sub-Saharan Africa, there is already a system in place to disseminate this ‘tool’. The tool is easy to build, but what takes time is validation – we need to ensure that the right content, the right partners and the right languages are available.
Extreme poverty
While the cost of the Ukanzizi pilot project was covered by the NGO “Opportunity”, which is funded by partners such as Cisco Systems and Mastercard, it is now seeking funding for a wider rollout in Malawi and elsewhere in the world. “Opportunity” said the funding will come from corporate and philanthropic donors.
Smallholder farmers often live in extreme poverty and there are hundreds of millions around the world who need access to best practices.
Smallholder farmers currently produce only 20% to 30% of their potential yield. Agriculture needs to double productivity in grain yields and triple vegetable yields to feed the planet by 2050, so this is a critical time to ensure smallholder farmers become the best producers they can be.
The team started with Malawi, where the country’s government was open to working with aid agencies such as the World Food Program and seeking help for the population that depended on agriculture – especially after Cyclone Freddie.
Although the government has agricultural extension agencies, there just aren’t enough of them. Nationally, the ratio was a factor for 2,500 to 3,000 rural households, against the recommended per 500 to 700 households. Around the capital there are just 341 advisers for nearly a quarter of a million farmers. In fact, visiting each farmer “is a tall order”.
These metrics have made the Uangizi app attractive.
Challenges
The farmer support agent suggested we just take a photo and send it in so the problem the pigs were having would show up. They did, and the app explained in detail what the pig was suffering from and what medicine the farmers should buy. If this application had been in place last year, perhaps so many of the farmers’ pigs would not have been lost.
Granted, there are challenges. In Ndodo, there is only one smartphone for 150 villagers. Also, connectivity is not universal in Malawi and data costs – while among the lowest in Africa – are still too high for farmers in one of the continent’s poorest nations.
Cell phone network around this place is poor. Besides this, the cost of data bundle is high for small scale farmers whose incomes are not fixed.
About 2 million Malawians have access to the internet and there are 12 million registered sim cards, said Moses Kunkuyu, the country’s information minister. However, access to smart devices is limited, he said. That said, farmers can access WhatsApp on some of the most basic phones.
Farmers who participated in the AI application trial say it saves them time and money.
It is faster to work in the application. In the past farmers could wait days for farm workers to come and deal with whatever problems they had on their farms. This is no longer the case. With a click of a button they have all the information they need.
