Pick Up
960. Can ChatGPT Adequately Answer Questions from African Rice Farmers?
960. Can ChatGPT Adequately Answer Questions from African Rice Farmers?
ChatGPT (Chat Generative Pre-trained Transformer), a conversational computer program using artificial intelligence (AI), has attracted considerable attention in recent years for its ability to learn from large amounts of text data and generate natural sentences.
AI such as ChatGPT has shown great promise in solving problems in the business and legal fields, as it can learn a variety of information and respond immediately to various consultations. If AI can be used to solve farmers' problems in the agricultural sector, it is expected to be used in regions and by farmers with limited access to agricultural extension services.
In sub-Saharan Africa, where agricultural extension workers are severely limited, the question arises as to whether AI can adequately assist farmers with their inquiries. To answer this question, Dr. Kazuki Saito and a team of researchers at the International Rice Research Institute (IRRI) conducted a study to evaluate ChatGPT's responses to queries from irrigated rice farmers in Nigeria. Their findings, published in Scientific Reports, shed light on this question.
The study involved formulating 32 questions related to irrigated rice production in Kano State, Nigeria, and soliciting responses from six agricultural extension agents. These responses, along with those generated by ChatGPT, were evaluated by four local experts, focusing on their suitability as recommended practices for rice production in the region. Notably, the responses generated by ChatGPT generally received higher ratings than those generated by the agricultural extension agents. However, instances where ChatGPT fell short were questions requiring specific numerical data, such as planting time, seed rate, and fertilizer application rate and timing.
The study concludes that while ChatGPT shows promise as a supplement to agricultural extension services, it's imperative that AI systems incorporate localized crop management details, including site-specific agronomic practices for input rates and timing. This would ensure the provision of practical and tailored information to farmers.
Reference
Ibrahim, A., Senthilkumar, K. & Saito, K. Evaluating responses by ChatGPT to farmers’ questions on irrigated lowland rice cultivation in Nigeria. Sci Rep 14, 3407 (2024). https://doi.org/10.1038/s41598-024-53916-1
Contributor: IIYAMA Miyuki (Information Program)