EMBRACING SKYNET: LEARNING AND DEPLOYING AI PLATFORMS IN THE CLASSROOM AND BEYOND

Currently, the educational community devotes more time on combating student use of artificial intelligence (AI) to prevent academic misconduct than embracing a new tool that could become part of professional and educational geoscience skill sets. Advanced computational approaches including AI may provide industry with opportunities to streamline projects by "outsourcing" repetitive tasks to AI, while educators can use AI to evaluate student work more equitably. We will present (1) new workflows that include in-class exercises for teaching students to leverage AI to create useful scientific products, and (2) in development platforms to aid in quantitative grading of student deliverables, including graphic products such as bedrock geologic maps, structure contour and isopach maps, and subsurface cross sections. In-class exercises include using a basic neural-network to create a geologic map from data points and digital elevation data in ArcGIS. The AI grading platform is based on the NVIDIA Jetson Nano system running a custom image similarity analysis protocol, providing educators with a "likeness" score of student projects with an educator-developed key.