Buying into the whole System and not a Just an App
In the year 2035, Pravesh, a smallholder farmer in Karnataka, watched as the agriculture industry in India underwent a profound transformation driven by AI and precision agriculture. It was a technological revolution that offered the potential for greater profits, by making agriculture more efficient – where agricultural outputs were determined by market analytics, but it also came at a cost, particularly for smallholder farmers like Pravesh.
Pravesh realised that to avail the benefits of these new technologies and compete with the high-yield crops that others in his area were cultivating, he needed to invest in modern tools and practices. This transition, however, came with a hefty price tag. He had to take loans to gain access to the internet and acquire a mobile phone, which had become essential for managing his farm. The advisories he received also told him to purchase high-yield patented seed varieties, the cost of which was significantly higher than traditional seeds.
Taking the advice of AI enabled crop advisory platforms, Pravesh began to grow high yielding varieties of paddy, abandoning his earlier practice of rotating between sorghum and pigeon peas. As soil health declined, he began to apply more fertilisers based on the inputs provided by the AI companies. While the AI companies were good at ascertaining how much of which fertilisers and medicines his soil and crops required, he was unable to understand the rapid decline in soil health.
As AI advisory platforms took over, the earlier plant doctors and Krishi Vikas Kendra extension workers who earlier frequented his village came less and less. No recourse to actual agricultural knowledge expansion, he became dependent on the minute to minute advisories provided by the numerous AI apps on his phone. The applications that provided farm advisories had also rolled out credit and financial incentives to adhere to the farming schedule. These platforms collected all kinds of data about his purchases, his farming activities, and other details. Pravesh now had to buy the inputs that were promoted and sold on credit (at high interest rates) by the AI apps on his phone, follow the ‘advice’ of the chatbot to qualify for crop insurance (which he must pay for), sell their crops to the company (at a non-negotiable price), and receive payments on a digital money app (for which there is a fee). Any missteps could affect his credit worthiness and access to finance and markets. At times he couldn’t stick to the schedule, he became worried that failure to adhere to the steps would impact his crops and ability to get credit.
The pressure on farmers like Pravesh to adopt precision agriculture and data-driven practices had created new dependencies. Smallholder farmers, already struggling with limited access to resources, found themselves at a disadvantage. As they accumulated debts to finance their technological investments, they became increasingly vulnerable to fluctuations in crop prices and external factors like climate change. Their rich wealthy neighbours and the dominant castes in the village often pressured them to sell their land.
Ultimately, Pravesh’s life was no better than before: he was still locked into a system of dependency on big corporations for seeds, pesticides and fertilisers. Newer dependencies also emerged as he became habituated to AI advisories, which led to his deskilling and the loss of traditional knowledge practices. He slowly began to grasp that selling off his farm was his best recourse, as AI led farm mechanisation was simply not feasible for farms as small as his. Under pressure from the wealthier and more dominant caste members in his village, he sold off his farm for meagre income, just enough to cover his debts. The future for Pravesh remains unclear, not knowing how he or his sons will adapt to this new reality.
Pravesh’s story is emblematic of the complex trade-offs and dependencies that smallholder farmers in India face in the pursuit of higher agricultural productivity. On one hand, AI advisories help plug in existing gaps in knowledge, but on the other, these applications also run the risk of de-skilling farmers, and making them dependent on AI. Secondly, current research on the role of big tech companies in the agriculture landscape in India also points to the fact that these advisories may not be neutral – companies might ring fence farmers into specific path dependencies such as specific seeds, and pesticides. This will also have the knock-on effect of intensifying monopolies in the agricultural sector, as well as carry the potential to create ecological damage, and a move away from sustainable practices.
In part 1 of this series, we explored the gendered dimensions of AI’s implications for Indian agriculture. In part 3, we will explore the potential pathways for action inorder to foster equitable and sustainable AI futures in Indian agriculture.
