This study evaluates the performance of OpenAI's ChatGPT AI chatbot in the accounting field. Results showed students had higher accuracy than the AI, but it was competent in True/False and Multiple Choice questions. Over 186 universities from 14 countries joined the research to better understand ChatGPT's impact on the education system. OpenAI is a leader in developing AI-based technology, and its groundbreaking work has the potential to revolutionize the way we learn and teach.
. OpenAI's ChatGPT technology has amazed the world with its capabilities since its launch in November 2020. In a study conducted by Brigham Young University and published in the American Accounting Association journal, its potential was tested with accounting paperwork. ChatGPT scored an impressive 47.7%, but still lagged behind the students average of 76.7%. ChatGPT was tested on tax, audit, management and financial evaluation and even exceeded the student average for accounting data systems and auditing with 11.3%. Despite these successes, it failed to overcome mathematics-oriented tasks. Professor David Wood from Brigham Young University has expressed concern with it being used for cheating, but also believes it can improve the learning progress of students and faculty in educational institutions.
This study by David Wood and 327 other researchers from 186 universities compared the performance of OpenAI's AI chatbot ChatGPT and university students in accounting exams. Results showed that students outperformed the chatbot, sometimes outperforming it by as much as 11.3%. The knowledge bank of the chatbot was found to be impressive, sometimes inventing its own references and making mathematical errors. Researchers are hopeful that the AI technology is the future of teach and learning.
This study conducted by Brigham Young University (BYU) researchers in tandem with 186 universities from 14 countries, found that OpenAI chatbot ChatGPT wasn't quite a match for the exam, but "impressive" nonetheless. It had a harder time solving tax, finance and managerial question, but still provided explanations and references, even if wrong and often made-up. With its results, possible classroom implications for AI models are discussed.
Discover the revolutionary AI-driven digital agent, ChatGPT, built to address complex accounting exam questions. Human students outperformed the AI bot with average score of 76.7% compared to 47.4%. Learn how AI can still be improved to bridge the gap between AI and human interaction.
Explore the evolution of tech policy from Obama's optimism to Harris's vision at the Democratic National Convention. What's next for Democrats in tech?