EcoGPT: Reducing AI's Impact Across the Embry-Riddle Campus
Presentation Type
Poster Presentation
In Person or Zoom Presentation
In-Person
Campus
Daytona Beach
Status
Student
Student Year and Major
Senior, Meteorology Major
Presentation Description/Abstract
Artificial intelligence (AI) is a recent innovation widely used as a tool among students. According to Forbes, 90% of students pursuing higher education have used AI academically, indicating its broad reach across the college campuses. However, since the widespread introduction of AI, MIT Technology Review reports that data centers (representing the energy source for AI) have nearly doubled in their energy consumption compared to the US average. College students are a large demographic in this energy usage, and there must be ways to reduce AI’s carbon footprint across college campuses. Feeding parameters to individual AI models to reduce unnecessary characters, graphics, and other high-energy outputs can be an option to reduce AI’s environmental impact. This study used an EcoLogits Python Program to assess the energy impact of Chat GPT to determine energy outputs on specific chats. A chat with no environmental parameters provided a control element to compare with the revised energy efficient version. By integrating these findings into an AI usage component of University 101, Embry-Riddle Aeronautical University (ERAU) can advocate for AI sustainability across college campuses. Using AI will continue to create a larger energy impact, but consumers at ERAU can choose to reduce this impact through a sustainable solution outlined in this research.
Keywords
Artificial Intelligence, College, Higher Education, Chat GPT, Microsoft Copilot, Data Centers
EcoGPT: Reducing AI's Impact Across the Embry-Riddle Campus
Artificial intelligence (AI) is a recent innovation widely used as a tool among students. According to Forbes, 90% of students pursuing higher education have used AI academically, indicating its broad reach across the college campuses. However, since the widespread introduction of AI, MIT Technology Review reports that data centers (representing the energy source for AI) have nearly doubled in their energy consumption compared to the US average. College students are a large demographic in this energy usage, and there must be ways to reduce AI’s carbon footprint across college campuses. Feeding parameters to individual AI models to reduce unnecessary characters, graphics, and other high-energy outputs can be an option to reduce AI’s environmental impact. This study used an EcoLogits Python Program to assess the energy impact of Chat GPT to determine energy outputs on specific chats. A chat with no environmental parameters provided a control element to compare with the revised energy efficient version. By integrating these findings into an AI usage component of University 101, Embry-Riddle Aeronautical University (ERAU) can advocate for AI sustainability across college campuses. Using AI will continue to create a larger energy impact, but consumers at ERAU can choose to reduce this impact through a sustainable solution outlined in this research.