Extending the Value of User Feedback Infinitely: A Look into the World of Entropi; Bridging the Gap: The Initiative Towards Ethical AI Development

Author:
Vittal, Aneesh, School of Engineering and Applied Science, University of Virginia

Advisors:
Norton, Peter, EN-Engineering and Society, University of Virginia
Vrugtman, Rosanne, EN-Comp Science Dept, University of Virginia

Abstract:

Artificial Intelligence (AI) offers extraordinary new possibilities with extraordinary risks. As we stand at a pivotal moment in the growth of these products, how does development of AI affect people’s sense of productivity and usefulness?

Customer feedback is valuable to businesses, but collecting and analyzing it is labor intensive and expensive. Using AI and large language models (LLMs), Entropi lets companies configure custom interviewer agents that represent them in a text-based interview with a user. With the user profiles that Entropi derives from customer responses, client companies can simulate future interviews on a new topic. The interviewer questions an agent constructed from the user profile, giving companies long-term use of the user’s original feedback. By combining
the advantages of live interviews and asynchronous surveys, Entropi offers extended feedback and analysis, saving businesses time and money

The recent proliferation of AI systems, especially large language models, has divided AI optimists from skeptics. The skeptics themselves are divided as well. Some warn that unregulated AI may constitute a future existential threat to humanity. Others, however, argue that AI has already been amplifying destructive biases inherited from training data. These divisions complicate regulatory responses to AI by dividing proponents of AI regulation into mutually hostile camps.

Degree:
BS (Bachelor of Science)

Keywords:
Artificial Intelligence, AI, Customer Feedback, User Research, Generative Agents