In online marketplaces, sellers can rely on alternative mechanisms to signal their quality when they lack rich transaction histories. Using scraped data on GPU sales from eBay, we find that certain image signals can substitute for reputation to increase conversion rates amongst sellers with less than 100% positive reputation, and conditional on making a sale, can realize a 5% price premium on average. However, the effects are only significant for less reputable sellers.
RESEARCH STATEMENT
My research interests are broadly the application, innovation and impact of AI on platforms and in the future of work. In particular, I am interested in:
- Computational creativity
- Generative AI
- Human-machine collaborative systems
- Economics of unstructured data
PUBLICATIONS
Coming soon!
CONFERENCES
-
(INVITED) Eric Zhou and Dokyun Lee. “Generative AI, Human Creativity, and Art.” Workshop on Data Science in Phoenix, Arizona.
-
(INVITED) Eric Zhou and Dokyun Lee. “Generative AI, Human Creativity, and Art.” Wharton Business & Generative AI Workshop in San Francisco, California.
-
(Presented by DK) Eric Zhou and Dokyun Lee. “Generative AI, Human Creativity, and Art.” Choice Symposium 2023 in Fontainebleau, France.
-
(INVITED) Eric Zhou and Dokyun Lee. “Generative AI, Human Creativity, and Art.” 2023 INFORMS Annual Meeting in Phoenix, Arizona.
-
Eric Zhou, Xiang Hui, and Dokyun Lee. “Economics of Image-Based Seller Quality Signals.” WISE 2022, Copenhagen, Denmark.
WISE 2022 Best Student Paper Finalist
WORKS IN PROGRESS
-
Eric Zhou, Xiang Hui, and Dokyun Lee. “Economic Value of Image-Based Seller Quality Signals.”
-
Eric Zhou and Dokyun Lee. “Generative AI, Human Creativity, and Art.”
Recent artificial intelligence (AI) tools have demonstrated their ability to produce outputs that society traditionally considers "creative". One such system is text-to-image generative AI, which automates humans’ execution to generate high-quality digital artworks. Utilizing a dataset of over 4 million artworks from more than 50,000 unique users, our research shows that text-to-image AI substantially enhances human creative productivity by 25% and doubles the perceived worth of artworks. While peak artwork novelty increases over time, average novelty declines, suggesting an expanding but inefficient creative space. However, AI-assisted artists who can produce the most novel ideas, regardless of prior ideation and/or mechanical skill, produce artworks that are evaluated more favorably by their peers. The results suggest that ideation ability is a necessary skill in the text-to-image process, thus giving rise to “generative synesthesia” – a blending of human thinking and perception that enables certain people to push the creative frontier with generative AI.
Invited to 2023 INFORMS Annual Meeting in Phoenix, Arizona
Invited to Choice Symposium 2023 in Fontainebleau, France (presented by DK)
Invited to Wharton Business & Generative AI Workshop in San Francisco, California
Invited to Workshop on Data Science in Phoenix, Arizona
-
Dokyun Lee and Eric Zhou. New Generative AI Project Coming Soon!
Coming soon!
-
Eric Zhou and Dokyun Lee. Another new Generative AI Project Coming Soon!
Coming soon!