Scene 1 (0s)
[image]. The Potential of AI Extension Agents to Support Women Home Gardeners in Ghana: An Human–Computer Interaction for Development (HCI4D)-Grounded Assessment.
Scene 2 (26s)
Foundation for This Study. “Leveraging social media for eco-education: home gardening for climate resilience and food security”.
Scene 3 (56s)
Literature Snapshot. Women in Gardening Primary managers of household food security. Vital role in preserving agrobiodiversity and seeds..
Scene 4 (1m 15s)
Problem Context and Equity Gap. Extension Crisis: High farmer-to-agent ratio (approx. 1,500:1) leads to overstretched public advisory services. Gender Disparity: Women receive fewer extension visits due to time, poverty, and mobility constraints. Digital Access Inequalities: Women are substantially less likely than men to own smartphones and access mobile data, restricted by cost barriers, lower literacy, and social norms. Equity Gap: Limited access to timely, gender-responsive agronomic advice hinders productivity..
Scene 5 (1m 39s)
Emergence of AI in African Agriculture. Across Africa, AI advisory services are expanding rapidly – conversational chatbots, voice assistants, remote sensing systems, image recognition that uses machine learning for plant disease diagnosis, predictive analytics for weather forecasting, local language voice interfaces, and generate valuable market information, among other uses..
Scene 6 (2m 7s)
Problem Statement, Research Question, and Objectives.
Scene 7 (2m 32s)
HCI4D Theoretical Framework. [image] odi. Core Principles: Contextual Fit Participatory Design Inclusivity & Accessibility Empowerment Sustainability Human-in-the-loop Analytical framework for: technical design, social integration, institutional embedding. Critical in low-resource contexts to ensure technology design aligns with local constraints, cultures, and specific user needs and capabilities..
Scene 8 (2m 49s)
Opportunities of AI Extension Agents. [image] Persoralized Advice hyper-local weather alerts Pest &sease fignosis Voice•frst multina access.
Scene 9 (3m 12s)
Risks and Design Challenges. Exclusion & Access: Risk of widening the divide due to high device/data costs. (Mitigation: Offline/USSD support). Unreliable Outputs: Hallucinations or incorrect agronomic advice. (Mitigation: Human-in-the-loop validation). Dataset Misalignment: Models trained on Western data may ignore local crops. (Mitigation: Contextual local data). Privacy & Surveillance: Concerns over data ownership and gardener privacy. (Mitigation: Transparent consent). Knowledge Erosion: Loss of indigenous wisdom and community learning. (Mitigation: Integration with community forums).
Scene 10 (3m 38s)
HCI4D Grounded Design Implications. Co-design directly with women home gardeners. Support multimodal: voice-first, offline, and USSD access. Use local-language and culturally resonant interactions. Build and utilize Ghana-relevant agricultural datasets. Establish clear transparency and community feedback loops. Ensure low-cost access and institutional sustainability..
Scene 11 (3m 55s)
Four Pillar Implementation Model. [image] CLIMATE RESILIENCE FOOD SECURITY.
Scene 12 (4m 18s)
[image]. Conclusion AI Extension Agents can significantly expand advisory reach to underserved populations. However, an equity-centered HCI4D design approach is non-negotiable for success. A hybrid community-institution ecosystem remains the path to durable impact. Questions & Discussion.