The Future of AI at UT
By: Rohit Gunti
As one of the top Universities in the nation (no bias here), UT should be dreaming big about the possibilities that could take student experience to it to the rocky top. With the AI Tennessee initiative led by UT, the sky truly is the limit. This initiative is focused on Smart manufacturing, climate smart agriculture and forestry, precision health and environment, future mobility, and AI for science. It’s mission is clear: advance cutting edge research and creativity across all disciplines at their intersections with AI.
UT is already working on developing student skills for them to get placed in AI-enabled jobs of today and the future. However, there should be an increased focus on AI in the classroom that helps students’ academic progress.
Nowadays, the main requirements for a healthy academic career for any student, whether they are international, national, local, or out-of-state, are digital/online resources, as well as the necessities for living like food, housing, and an accessible commutes to college. While UT students have access to AI-based food delivery robots on campus and digital resources to help us study, housing and transportation could be areas where AI tools may be incorporated into practice. Expensive housing and low vacancies in Knoxville could potentially cause more students to leave the campus, just like it happened with Arabella Sarver. In fact, Sprinklr, our AI-driven social media analytics tool, tells us that the “housing crisis” in Knoxville has been mentioned over 450 times since the beginning of the school year (August 23rd, 2023) times across social channels. The topic has reached over 130K social feeds with a spike in conversation near the beginning of the classes:
This is where AI comes in. If more spaces are available closer to campus at reasonable prices and we begin to use AI technology to help students find housing when they need, the game could be changed forever. Instead of relying solely on humans to answer calls and queries for limited hours in housing administration, an AI-bot could be deployed to answer student questions the rest of the time, or set up virtual tours, give recommendations, and perform neighborhood analysis. Additionally, smart home automation should be encouraged in every apartment to make student life and billing more effective.
When it comes to transportation, instead of relying too much on Google Maps, it would be easy to train a recurrent neural network algorithm that can learn inputs like time of the day, day of the week, previous bus arrival times and predict bus arrival. And just like that, we have several models like the transformer model used in ChatGPT that can be trained and deployed through a mobile app or website. The current trends among commuters at UT include themes of “parking and UTK” have reached over 5.3 Million social feeds times across social channels since the start of the school year with the net sentiment score graphic posted below:
UT should train and evaluate as many open-source AI-based IR systems as we can and take observations to build the newer, successful systems to help student and other stakeholders. This entails building more labs and teams that support this research with the focus on helping UT students. With AI, the possibilities are endless, and UT should be at the forefront of this exciting technology.