Working Paper

The Creativity Contest: Human Versus AI Chatbots in the Generation of Novel Ideas

Artificial intelligence (AI) has made substantial progress, but whether AI can be creative is not well understood. This study compares the ability to create new ideas of two generative AI chatbots -- ChatGPT and Bard – to a large and representative population of adult humans. Humans rate ideas generated by AI chatbots as more creative than those created by other humans. ChatGPT is substantially more creative than humans, while Bard lags behind. Competition from AI does not reduce human creativity much. Augmenting humans with AI improves human creativity, but ChatGPT alone is rated as more creative than human augmented with AI. Humans assign lower creativity scores to ideas they believe to be AI-generated, even though they cannot discriminate well between ideas created by AI or other humans. (slides on demand)

Work in progress

Assessing the Impact of Online Rating Systems on Consumer Welfare in Differentiated Markets

This study investigates the effectiveness of online rating systems in aiding consumer product choices, focusing on vertically and horizontally differentiated markets. By conducting an online experiment, it explores how rating systems, particularly those with added features like filtering and freezing periods, impact consumer welfare. Participants, tasked with selecting and rating quizzes in artificial marketplaces, provide insights into how ratings influence decision-making and welfare in different market structures. The study aims to determine whether rating systems enhance welfare and if modifications to these systems can further optimize consumer outcomes in varied market contexts.