The Potential of Neural Network Models on the Example of ChatGPT: Opportunities, Limitations and Application in the Analysis of Foreign Trade

Abstract

Contemporary neural network models are facing a rapid growth of application in a wide range of applied tasks. Consequently, this topic is popular in public discourse, but is oversimplifi ed – the attention is given to individual products, such as ChatGPT. Th us, there is a certain contradiction – it is not the neural networks that are gaining popularity, but just one of their kin. In addition to that the potential for using neural networks is signifi cantly exaggerated in public discourse, just as their importance in generating new knowledge. Based on the identifi ed issues, the study aims to prove the neural networks’ viability as a tool of applied analysis and a fi nancial product with full respect to their potential and usage limitations in the analysis of international relations. Th e following tasks have been solved to achieve this goal: the possibilities and limitations of using neural network models have been identifi ed; the public discourse on the relevant topic has been analyzed; the potential of ChatGPT as a successful commercial project has been highlighted; the potential for the analysis and management of foreign economic relations is identifi ed. Th e most signifi cant results of the study include: proof of the applied signifi cance of the development and use of neural network models; revelation of the trend nature of interest to the individual products based on neural networks; proof that ChatGPT is a powerful driver of the transformation of the global high-tech market.