Abstract for: A Review of Five Key Articles on “How to use LLM in System dynamics (SD) modeling”
System Dynamics modeling process can be complex and time-consuming. Therefore, having a powerful virtual assistant can be a great help to the SD modelers. Generative Artificial Intelligence (AI) and Large Language Models (LLMs), e.g., tools like ChatGPT, have received widespread recognition from researchers[1]. This is also true for system dynamics modelers. In this study, we will review five papers that have utilized LLM for creating System Dynamics (SD) modeling. - How can Large Language Models (LLMs) contribute to System Dynamics modeling? - Do Large Language Models (LLMs) from different companies perform equally in System Dynamics modeling? - Should the interaction with all Large Language Models (LLMs) from different companies be the same for System Dynamics (SD) modeling? - Is the use of Large Language Models (LLMs) effective in System Dynamics (SD) modeling? Based on our study of these five articles, Different LLMs can be acceptable SD models provided that humans identify and apply optimal interactions with each LLMs. it seems that in the future, an important role of humans in SD modeling will be to discover and apply optimal interactions. Also to create highly accurate SD models, we can compare SD models created by LLM from different companies. In general, the steps of these five articles that reviewed in our study can be summarized in the following three steps: 1. Selecting models of LLMs 2. Interaction with models of LLMs 3. Validation of the AI-generated SD models (for example by Comparing it with human-generated SD models) monica- to correct the rules of writing in english