AI Seed Fund Recipients

This project will work to increase students' knowledge of how learning and memory works by having them skeptically interrogate claims of LLMs about the neural substrates of learning and memory, with feedback provided from senior students about different ways that users can reveal weaknesses, errors, or holes in the scientific information that LLM systems are providing in response to prompts

This project will help teachers of discipline-specific (CL2) writing courses become more comfortable teaching with GAI. Participants will incorporate Gen-AI in learning modules to help students explore how these tools influence the writing process.

This project will explore the use of AI in elucidating patterns of instructor course evaluation data across an institution. The project aims to develop an AI that can perform a sentiment analysis on the qualitative student comments in course evaluations. This AI-generated metric will then be used to investigate patterns of course evaluation data across an institution.

This project aims to cultivate an infrastructure of sustained collaboration that leverages expertise from language and area studies, linguistics, computer science, and the health sciences, in service of advancing the ethical development of a Spanish LLM.

This project explored the creation of open source code, sample projects, and a collection of relevant videos and texts that support artists and designers as well as other students at UB in their investigations of ethical and cultural edge conditions of AI use.

This project aimed to create pedagogical materials for teaching presentational writing skills in French and Spanish language courses. In addition, it set out to demonstrate to TAs, adjuncts and students in Romance Languages and Literatures how AI can be helpful as well as detrimental.

For the Game Studies program at UB, this project worked to explore how generative AI tools can be incorporated across courses to be used by students in creating the raw materials of projects. This follows expected future trends in the professional game industry.

This project will introduce Troubleshooting Problem-Solving Instruction assisted with AI (TPS+AI) to improve troubleshooting skills and performance in a circuits course. A controlled study will assess the impact of TPS+AI versus traditional instruction.

This project will integrate LLM-Enhanced, machine-assisted proof systems, such as Lean, into the Algorithm course to enrich curriculum offerings, streamline proof decomposition and verification, and automate the visualization of algorithms and proofs.

This project will develop AnatomyGPT, a new learning tool based on an open-sourced LLM and fine-tuned on our expert-crafted Gross Anatomy curriculum materials. We will do so with the goal of creating a first-in-class AI learning assistant for the anatomy lab.

This project aimed to prepare students to be active and continuing learners throughout their careers by utilizing Large Language Models’ ability to search, summarize, and compare literature; and perform instructional tasks (Natural Language Programming).

This project will seek to integrate AI into fixed prosthodontics to transform teaching with interactive lectures and AI-driven personalized assessments. It will do so by replacing traditional quizzes and exams to enhance student engagement, improve outcomes, and personalize learning.

This project will explore how AI can use machine learning for alignment algorithms. Alignment of STL or PLY files are important part of dental procedures. New VR googles are making it possible to see 3D files in real time with different designs and changes. Alignment of different files are possible using virtual reality googles and can be seen simultaneously with different designs reflecting different scenarios that is making it possible to progress the treatment and rehabilitation of patients.

Periodontal and Dental Technologies was a new course designed to expose postgraduate students to state-of-the-art dentistry. Our goal for this project was to incorporate advances in AI through consultation with experts who visited our facilities and lectures.

This project explores online simulations for health professions students to address vaccine hesitancy. The aim of this project will be to develop online simulations, with adaptive, personalized feedback for health professions students to be trained in addressing such hesitancy.

This project, titled EASE, leveraged local Large Language Models (LLMs) to elevate faculty and student engagement with AI. The platform was designed to provide a suite of AI-powered micro-applications with a focus on enhancing instructor efficiency and student learning.

This project sought to demystify the inner workings of generative AI tools such as ChatGPT. Here, we sought to produce a graphically rich tool to help people visualize how these inhuman algorithms produce close-to-human results.

This experimental project aimed to reinvent a graduate seminar course with high-touch and high-tech education methods, utilizing the Socratic teaching method and ChatGPT technology support together to improve students' critical thinking and creativity.

The project aimed to revolutionize architectural design and education by integrating generative AI technologies such as ChatGPT, Midjourney, and NeRF modeling into novel ways of addressing the architectural design process.

Pertaining to students' self-directed and ambitious hypothetical design projects, Chat GPT+ serves as a critical evaluator and idea articulator. Correspondingly, Midjourney acts as an AI-driven aesthetic generator. As such, this project explored how, together, these tools may accelerate the creative design process.