June 16, 2023
Assistant Professor Morgan Frank (Department of Informatics and Networked Systems) collaborated with computer scientists, legal scholars, economists, designers, cultural researchers, and artists on a new Perspective article published in the prestigious Science magazine describing misconceptions and open questions about generative AI, including ChatGPT and ImageNet, and the future of art.
The article highlights new legal, cultural, and economic implications brought on by these AI tools. For example, AI that is trained on existing artistic media and then produces new original media challenges traditional notions of copyright and ownership. Economically, these tools may automate artistic jobs (e.g., for graphic designers or writers), although historical examples suggest otherwise. Culturally, generative AI has the potential to mass-produce content that could become training data for future AI systems thus producing a feedback loop with the potential to dampen artistic diversity.
The published article, which can be read here, is accompanied by a longer white paper that can be found here. This work is also featured in The Conversation, MIT News, and ZDNET.
Pictured left: an AI-generated image created by Frank in Dream Studio. Pictured right: an AI-generated image of a face created by study co-author Memo Akten.
