AI Generation Time Machine
Professor: Marti Hearst
Description: AI generation systems as embodied in large multimodal (language) models are constantly changing in their abilities. The goal of this project is to create a digital “time machine” that creates a record showing the quality of the output that different models generate at different points of time. Of special interest are models that produce images and visualizations, to see how these change over time. The main deliverable will be a website with a highly usable, searchable dataset of LMM-generated output. Students will gain skills both in using LMMs and creating an interactive website.
Preferred Qualifications: Ideally has taken an HCI course and has experience with python flask. Python experience is a must