Publication Details
Abstract
Technologies in AI, namely Machine Learning (ML) and Neural Networks have begun to influence various economic fields such as macroeconomics and international trade. The shift from symbolic reasoning to more exploitative methods have paved new paths for the field to tread on, the most primed, perhaps, in economic analysis in the form of predictions and policies. The integration of AI into economics has been slow, working hand-in-hand with the developments of computational power and Big Data. The historical evolution of AI in economics has transitioned from expert systems to ML methods, which are now increasingly more about prediction and less about shape the world through optimal solutions by reusable patterns. However, whilst AI shows great potential, its take-up in economic research, particularly within the fields of macroeconomics and international trade, has lagged behind other fields such as the sciences. We have a very little understanding of how to better integrate AI with traditional economic mode of operation for more equity, effective decision making. Abstract Our approach in this study uses the scientometric methodology to examine the diffusion of AI methods adoption in macroeconomics and international trade. It gives theoretical perspectives on institutional roles, regional and academic structures driving this change, as well as empirical findings on changes made in the use of AI-augmented systems for doing economic modeling and planning. Advances in machine learning and associated concepts such as neural networks led to more promising applications of AI which started generating a lot of interest among other fields and these fields would start using these methods.