Abstract for: AI-Powered Inputs for Causal Loop Diagrams versus Human-Driven Inputs on Factors Affecting Electric Vehicle Adoption

Electric vehicles (EVs) have gained immense popularity, especially in highly industrialized nations that are keen on reducing their dependency on fossil fuels. Most of these vehicles have been adopted through regulatory measures, offers of subsidies, and/or a set of related measures that tend to tip the economic balance of EVs in favour of the users. South Africa, as a developing economy, faces unique challenges in adopting EVs. This qualitative study investigated the driving forces which impact EV adoption in south Africa through artificial intelligence (AI)-driven approach using multiple AI-engines. The results were then compared to those factors identified from workshop engagements (Human intelligence). This resulted in the development of a comprehensive set of variables that could be used to develop a causal loop diagram (CLD). Through the use of AI models, it was possible to identify other factors that stakeholders can be engaged on in upcoming workshops and which can be included in a CLD or system dynamics model. The problems noted with the AI modelling process are: The process is often automated and lacks human interaction, which may limit stakeholder buy-in and shared understanding; does not allow iterative refinement through discussion, incorporating lived experiences and tacit knowledge that AI may overlook and the output may still require expert interpretation to validate and make it actionable. The research was on comparing AI to HI