Is ChatGPT better at business consulting than an experienced human analyst? An experimental comparison of solutions to a strategic business problem Cover Image

Is ChatGPT better at business consulting than an experienced human analyst? An experimental comparison of solutions to a strategic business problem
Is ChatGPT better at business consulting than an experienced human analyst? An experimental comparison of solutions to a strategic business problem

Author(s): Łukasz Sulkowski, Dawid Jarco
Subject(s): Business Economy / Management, ICT Information and Communications Technologies
Published by: Wydawnictwo Naukowe Akademii WSB
Keywords: decision making; management; ChatGPT; AI; LLM; financial analysis;

Summary/Abstract: Given the level of continued technological progress, AI language models have been the subject of enormous me- dia attention. The popularity of such models skyrocketed in November 2022 with the introduction of the ChatGPT application. The impact of AI language models on areas such as research, healthcare and business is constantly be- ing evaluated. However, an issue in which there is a clear research gap is the use of ChatGPT for business deci- sion-making purposes. The aim of this article is, therefore, to investigate whether ChatGPT can be leveraged to make strategic business decisions. The methodology compares three paths towards a case for a strategic business deci- sion by means of a checklist-based score: one conducted by a human alone; the second taken purely by ChatGPT with as little human interaction as possible; and third, the centaur model, where a human actively interacts with ChatGPT and asks critical questions. The findings were twofold. Firstly, ChatGPT is still far from being considered a quality business consultant in itself. Secondly, however,AI turned out to be extremely helpful in terms of creating a decision-making path. Hence, humans may leverage the information provided by ChatGPT to gain helpful insights for better business decision making.

  • Issue Year: 11/2023
  • Issue No: 2
  • Page Range: 87-109
  • Page Count: 23
  • Language: English
Toggle Accessibility Mode