Application of ChatGPT in seepage-induced slope stability

Abstract

ChatGPT has been recently applied to many engineering fields including geotechnical engineering. This study investigates the chance of using ChatGPT for seepage-induced slope stability problems based on the grid and radius method. The Python code for solving seepage and slope analysis was first generated, followed by the coupling of those two analyses was performed using ChatGPT. In addition, the grid search method and three scenarios of optimization methods were applied to determine the best optimization method for the computational efficiency of the developed framework. The comparable factor of safety (maximum error of 1.86%) shown in this study compared to those obtained by commercial software demonstrated the feasibility of using ChatGPTgenerated code for seepage-induced slope stability analysis. Furthermore, the use of optimization techniques enabled up to a 70% reduction in calculation time, and the relatively easy implementation of optimization methods using ChatGPT implies that the appropriate prompts in ChatGPT can provide a wide range of applications in slope stability analysis.