Artificial intelligence (AI) is a hot topic that is raising a lot of discussion on how it will impact higher education. LSA Technology Services supports the appropriate use of AI across the college, and we are already helping instructors navigate best practices with this new exciting tool.
AI in the Classroom
LSA Technology Services Learning and Teaching Consultants (LTC) are assisting faculty with navigating the unknown territory educators find themselves in with the recent widespread availability of generative AI (GenAI) tools. Emily Ravenwood, Learning Technology Consulting Manager in LSA Technology Services, works to anticipate the impact AI will have in higher education. As a starting point, she recommends visiting U-M’s generative artificial intelligence website to learn more about the three AI tools available from the university. They include U-M GPT, U-M Maizey, and U-M GPT Toolkit. They are maintained by ITS and are accessible to people with many different levels of experience with AI. These tools are a secure alternative to public GenAI tools like ChatGPT, and are trained using responsibly sourced data.
LTC shares information on how to use AI in the classroom across a variety of formats. The team has put together Guidelines for Using Generative Artificial Intelligence. These recommendations include: state clearly in your syllabus what use of GenAI is acceptable for class (if any), and do not use GenAI for grading. LTC has also put together workshops regarding GenAI in teaching for discussing general use as well as tailored content for specific departments by request. Departments have requested a range of workshops, from AI basics about what it is and what it can do, to assignment redesign and finding creative ways to integrate AI into courses. Instructors are also asking how to communicate the importance of not using AI to cheat, as class assignments are supposed to benefit students and their learning process. Ravenwood says it helps to talk directly to the students and explain the purpose of the assignments. “Students are very good at that calculation — what will benefit me in my future and my career?”
Some faculty are already digging into the possibilities of U-M’s GenAI tools. Dr. Lynn Carpenter from the Earth and Environmental Sciences department, for example, is exploring how to use Maizey to create a study buddy with GenAI. Ravenwood reiterates that whatever faculty “want to do or worry about doing with GenAI, they can come and talk to LSA Technology Services about what this means to their class in particular.”
AI in Research
Luke Tracy, Research Computing Services Manager in LSA Technology Services, is also assisting in the organization’s efforts to support LSA with AI. Tracy’s team provides a wide range of support and tools for researchers.
Osman Khan, a professor in the LSA Digital Studies Institute and the Stamps School of Art and Design, created an innovative storytelling tool with LSA Technology Services’s support. Khan’s project challenges the bias in AI, which often relies on datasets developed by white Americans. His storytelling tool is fed data on South Asian and Middle Eastern immigration narratives. The tool is made to be used with children from those regions. The stories that make up its dataset makes it more relatable to its target audience and avoids built-in bias. Tracy’s team is helping get the models working and feeding the data. They are also ensuring the software is working so it can create its own stories after the fact.
Many of the AI projects that LSA Technology Services assists with are related to Geographic Information Systems (GIS). AI can be used to do simple and repetitive tasks that require some programming logic. An example is object identification, like a project from the Ecology and Evolutionary Biology department with pivot irrigation fields. The object identification used to be done manually, but now AI tools in the ArcGIS software makes light work of it. While this use of AI is not new, the university continues to find new ways to innovate.
Graduate student Alfredo Cabrera used it to create imagery for a music video. Criteria were fed into the engine to create imagery and the synchronization of the music with the generated imagery was done by AI.
Tracy believes there’s no telling where AI will take us. “Something like AI could infiltrate everything and have an effect and impact on every sector of society, and it’s hard to see where the limits are. But so many times in the past people have been unsure of where change will take us at the time, and it’s only in hindsight where we can see where there were limits.” But his team will rise to the occasion to “help researchers with and without a technology background explore that space.”
If you’re interested in learning more about how you can use AI in your research, classroom, or work, reach out to us at LSATechnologyServices@umich.edu and we will connect you with our highly skilled staff members to discuss your ideas.
