The Rise of AI in Agriculture
Adamjee Lukmanjee & Sons
Leading exporter of coconut products and spices, manufacturer of N-Joy and importer of cement and animal feed.
Artificial Intelligence seems to be integrating into all aspects of life and business. Several industries now rely on machines and servers to do critical work, and automation has transformed businesses in terms of their efficiency and output.
As the need for food and nutrients grows exponentially around the globe, so does the amount of pressure put on farmers and the agriculture industry. Farmers have now begun implementing AI technology into their farming practices to generate higher yields and reduce costs and resource wastage where possible.
So, what are the benefits of AI in Agriculture?
Generative AI in the agriculture industry can use the high volumes of unstructured data logged by farmers and industry experts to create structured, patterned trends and predictions regarding weather, soil quality, and the pressure of pest and disease.
Addressing the AI Disparity in the Global South
The benefits of AI in farming and agricultural practices are plentiful, however most of the benefits and case studies obtained for the list above come from the "Global North". Implementing AI and technological solutions in an industry that, to an extent, still thrives on some age old traditions, is tougher and less feasible. For example, some coconut growers in Sri Lanka still harvest their coconuts manually, by skilfully climbing up trees to pluck the coconut. The benefit of this is that the growers are able to inspect the coconuts before plucking them, and can identify any that are diseased or not suitable for picking. Others pluck the nuts using long poles with blades attached to the end of it. While this method is no doubt very labour intensive, the costs are minimal compared to using machinery and is therefore the preferred method.
One of the root causes of the 'AI Divide' is due to the structural and resource limitations that we experience. There is a requirement to install computing resources and storage capacity to handle large workloads and data mines. In the context of Sri Lanka, it is incredibly difficult to install such infrastructure in and around farmlands, let alone have the high bandwidth and network capabilities to handle such programming and processing. Reports show that training AI algorithms and running massive servers to do so comes with a big price tag. The cost being guaranteed to be several millions of dollars, smaller economies like Sri Lanka don't have the resources to implement such infrastructure at a large scale, let alone keep it going for the long run.
Additionally, the Foundation Models that are trained to store large data sets and thereafter adapt to downstream tasks for R&D and new use cases are largely controlled by a few companies in the US and China. These companies limit who can access the Foundation Models and what they are being used for.
领英推荐
There is also the question of whether many farmers in Sri Lanka and other countries of the Global South are willing to compromise on tradition and heritage for innovation and automation. The practices of agriculture are rooted in tradition and age old cultures, and farmers often feel a connection to their heritage through the land on which they farm. For example, the Coconut "Tree of Life" has deep rooted significance in many South Asian and South-East Asian cultures. In Indonesian culture, the tree represents fertility, replenishment, and a connection to the divine. There is therefore a debate that perhaps automating the farming process perhaps removes or dulls the cultural significance that people feel when harvesting.
Traditional practices of farming are also proved to be better for the environment than implementing modern automation practices, proving to be more sustainable for the future. In coconut plantations, the ecological footprint of harvesting is little to none. The product itself is known to be "zero waste" as all parts of the coconut are used in manufacture, and if not is put back into the ground as fertiliser and to enrich the soil. This is where the use of AI in the industry proves to be a problem. It is a well known fact that the ecological footprint of AI use is massive - the energy it takes to run servers alone is incredibly high and can cause a dent in both the economy and the environment.
In Conclusion...
There are no doubt many benefits to AI in agricultural practices. Farmers can use the powerful tools of AI to improve their positions in local and global markets, empowering them against big players in procurement and providing them with better leverage and knowledge. But, is it for everyone? Is this a practical and feasible solution for farmers and economies with fewer resources to work with? The costs associated with implementing AI tools in the agriculture industry in the Global South is perhaps too high for it to be a sustainable model of progress, at least for now. AI is yet to become a more commercial and accessible tool for everyone and until then it unfortunately means that a lot of the world may be left behind in the path to complete automation.
Resources: