Generational Visions of Urban Land and Green Infrastructure: An Extreme Case Sampling Approach to AI-Driven Participatory Planning with Children and Seniors

Public open spaces are vital for sustainable urban land management, yet design processes often marginalize demographic extremes like children and the elderly. While generative artificial intelligence (GenAI) emerges as a bridge to democratize public space design, its inclusivity across varying cognitive and digital literacy levels remains unexamined. This study deploys an “extreme case sampling” approach within a Design-Based Research (DBR) framework. Using the Human–AI Hybrid Planning Interface Model (HAIHPIM), we contrast empirical data from a rural child cohort (aged 6–13, n = 12) and an urban senior cohort (aged 60–96, n = 6) in Northern Germany, combining qualitative participant evaluations with ethnographic observations. Findings reveal a profound divergence in spatial imagination. Children exhibit high creative output but face cognitive barriers in prompt formulation, whereas seniors circumvent digital interface constraints through narrative spatial knowledge and historical reference mapping. Rather than a definitive tool for spatial justice, GenAI functions as an exploratory boundary object. It makes the informal spatial needs of marginalized groups discursively visible and highlights the indispensable role of human mediation in adaptive land governance.