Generative Artificial Intelligence for Sensory Trust and Resource Optimisation via Crossmodal Nudging

"Household food waste is a major environmental and economic challenge in the EU, with domestic kitchens being the largest source. Despite widespread public concern, a significant ""intention-action gap"" prevents most citizens from adopting more sustainable behaviours. Generative Artificial Intelligence (GAI) offers a promising solution by creating personalised recipes to minimise waste, but its effectiveness is critically undermined by a lack of consumer trust in its outputs.The GASTRO-X project will tackle this trust deficit by developing and validating a novel “crossmodal nudging” framework. Drawing on principles from gastrophysics, the project will systematically enhance AI-generated recipes with vivid sensory cues—using descriptive language, imagery, and sound—to make them feel more authentic and appealing. By increasing the perceived naturalness of the AI’s suggestions, we aim to strengthen user trust, accelerate the adoption of GAI-powered sustainability tools, and ultimately reduce food waste at the household level. The framework is designed to be highly adaptable and could be extended to promote other desirable behaviours, such as healthier eating.The project uses a mixed-methods approach, combining GAI performance benchmarking, large-scale online behavioural experiments, and the development of an open-source prototype. This research is fundamentally interdisciplinary, integrating insights from sensory science, behavioural economics, and human-computer interaction.The fellowship will be hosted at the University of Oxford’s world-leading Crossmodal Research Laboratory under the supervision of its director, Professor Charles Spence, a global pioneer in gastrophysics.This fellowship will provide the essential project leadership experience, high-impact training, and international network required to launch my independent academic career and secure a future tenure-track position."