A Pilot Study Investigating STEM Learners' Ability to Decipher AI-generated Video

Abstract

Artificial intelligence (AI) techniques such as Generative Neural Networks (GNNs) have resulted in remarkable breakthroughs such as the generation of hyper-realistic images, 3D geometries, and textual data. This work investigates the vulnerability of science, technology, engineering, and mathematics (STEM) learners to AI-generated misinformation in order to safeguard the public-availability of high-quality online STEM learning content. The COVID-19 pandemic has increased STEM learners' reliance on online learning content. Consequently, safeguarding the veracity of STEM learning content is critical to ensuring the safety and trust that both STEM educators and learners have in publicly-available STEM learning content. In this study, state-of-the-art AI algorithms are trained on a specific STEM context (i.e., climate change) using publicly-available data. STEM learners are then randomly presented with authentic and AI-manipulated STEM learning content and asked to judge the authenticity of the content. The authors introduce an approach that STEM educators can employ to understand correlations between STEM learning topics such as climate change, and students' susceptibility to AI-driven misinformation. The proposed approach has the potential to guide STEM educators as to the STEM topics that may be more difficult to teach (e.g., climate change), given students' susceptibility to AI-driven misinformation that promotes controversial viewpoints. In addition, the proposed approach may inform students themselves as to their susceptibility to AI-driven STEM misinformation so that they are more aware of AI's capabilities and how they could be utilized to alter their viewpoints on a STEM topic.