Design for Social Change: Mitigating Bias in AI-Assisted Design

This project is connected to the LUCID Framework, a project that imagines a scenario where the latent needs of myriad users are automatically elicited from social media, forums, and online reviews, and translated into new concept recommendations for designers. With the burgeoning development of AI, there is great potential to expand designers’ possibilities in eliciting and applying users’ needs.

This project highlights the call for awareness and responsibility, addressing the risk of transferring intrinsic biases of algorithmic translations into the design of products, services, and systems. This project will advance the fundamental understanding of if and how AI can augment the performance of designers in early-stage product development. We want to build on top of the state of the art by proposing a synthesized model of AI-designer collaboration in user need elicitation and their application in the design process, especially targeting the reduction of human-induced biases in programming NLP algorithms. We aim at testing out the models’ ability to align with human values and the benefits and shortcomings of implementing AI in the need-finding process for design.

View the CogSci 2024 conference paper.