| Sumario: | This study investigates the general farmer information ecosystem in Kenya. Furthermore, it examines the shortcomings of AI-generated agricultural advisories in Kenya, particularly in terms of gender inclusivity and cultural relevance. Despite recent advancements in AI, the advisories did not fully meet the diverse needs of smallholder farmers across Kiambu, Kakamega, Meru, and Nakuru counties. A total of 120 farmers participated in focus group discussions (FGDs), with equal representation of women and youth. The research aimed to validate AI-generated content by evaluating its clarity, relevance, trustworthiness, and inclusivity. Although the AI advisories scored reasonably well for clarity (1.58 out of 2), farmers expressed significant reservations about the cultural fit and gender inclusivity of the content. The lack of localised examples and culturally sensitive references was a recurring critique. Furthermore, while the AI advice was often clear, farmers were skeptical about the credibility of the source, particularly when delivered via SMS or other digital platforms. Trust in AI advisories remained low, with many farmers preferring traditional sources, such as extension officers or local radios, which they deemed more reliable. Gender and age disparities in access to and trust in digital platforms further compounded the problem, with women, especially older farmers, still relying heavily on face-to-face extension services. The findings reveal that AI-generated content, while technically sound, fails to fully align with farmers' realities and lacks the trust required for widespread adoption. The study underscores the need for AI systems to better integrate local knowledge, language, and gender-specific considerations to ensure they are truly effective in addressing the needs of Kenya’s farming communities.
|
|---|