Can You Tell Who's Funny?

Abstract

This study presents an innovative, multi-phase research design that explores the intersections of artificial intelligence, human creativity, and humour perception. At the core of our approach is the interactive platform https://3dhkilc89y35.manus.space, developed to collect jokes without collection bias. Contributors are invited to submit jokes while tagging the mode of creation — human-authored, AI-generated, or co-created. This playful yet rigorous method constitutes a novel pathway for assembling humour data that transcends cultural and technological boundaries.

In the second phase, the platform will serve as the basis for a survey experiment. Submitted jokes are presented randomly to participants, who are asked to evaluate their humour value and to attribute authorship (human, AI, or human-AI collaboration). This design allows us to examine perceptions of humour across modes of production and to test cultural frameworks of attribution and appreciation. The third phase integrates these findings through online focus group discussions with humour scholars and AI experts. These sessions will critically reflect on the survey results, situating them within broader debates on bias, stereotyping, and cultural context in humour studies and computational creativity.

Our research goals are twofold: first, to empirically assess how cultural frameworks shape humour perception in human-AI contexts; and second, to demonstrate how a playful, participatory platform can function as both a methodological innovation and a nudge towards rethinking humour as a transboundary practice. In doing so, we argue that digital comedy provides a fertile ground for understanding human-AI interaction, cultural negotiation, and the epistemology of laughter.

Keywords: Cross-cultural humor, Artificial intelligence, Computational comedy, Human-AI collaboration, Cultural bias