Researchers recently conducted a study to evaluate how well artificial intelligence chatbot ChatGPT fares on accounting exams. OpenAI’s tech was put to the test, since it is a game changer in several fields. Results found that while ChatGPT’s performance was impressive, students performed still better. On a 11.3 percent of questions posed, ChatGPT outscored the student average score, indicating it could be better suited for AIS and Auditing. However, ChatGPT had difficulty with questions requiring mathematical processes, especially in tax, financial and managerial assessments.
OpenAI is a technology company backed by Microsoft and is known for their newest AI chatbot product, GPT-4. This chatbot uses machine learning to generate natural language text and has passed demanding tests with high scores, such as the bar exam, AP exams and the GRE Verbal test. While GPT-4 is more advanced than ChatGPT, it still has work to do in the realm of accounting.
Lead study author David Wood, a professor in accounting at Brigham Young University, noted the importance of using new technology to improve the teaching and learning experience for both faculty and students alike. While ChatGPT performs well on true/false and multiple-choice questions, it had difficulty with short answer questions and higher order questions. It also proved to sometimes make up facts when providing references, which could prove problematic in an academic setting.
Overall, ChatGPT is still no match for humans when it comes to accounting, but GPT-4’s improves development has the potential to progress the industry on a new level. OpenAI has strong faith in the technology they have created and is expecting it to dramatically improve as it is adapted and developed.
