ChatGPT recently started making waves in December 2022, but experts are already touting it as the future of tech. However, academics are not so keen on welcoming this AI tool.

In fact, two of the most prestigious journals in the academic publishing realm, Science and Springer Nature, have introduced new editorial standards to exclude or severely limit the use of tools such as ChatGPT in writing academic studies.

However, a new study may serve as a devil's advocate for these concerns. It claims that it can serve as a research aide for academics and not a threat.

FRANCE-TECHNOLOGY-IT-AI
(Photo : LIONEL BONAVENTURE/AFP via Getty Images)
This picture taken on January 23, 2023 in Toulouse, southwestern France, shows screens displaying the logos of OpenAI and ChatGPT.

"Good Finance Journal"

Finance researchers Brian Lucey from Trinity College Dublin and Michael Dowling from Dublin University recently detailed their study's findings in an article published in The Conversation.

They claim that ChatGPT could be utilized to write a finance paper that would pass as an academic journal.

They first asked the AI tool to produce the four standard parts of a research study: the research problem, related literature, dataset, and testing and examination suggestions.

The researchers told ChatGPT that the output should be worthy of being published in a "good finance journal."

For another test, they pasted at least 200 abstracts of relevant research studies into the ChatGPT window and asked the program to take these into account in creating the four research stages.

They reviewed the answers made by the tool and gave suggestions for improvements to ensure that their expertise is still used in the endeavor. 

After that, the team requested a panel of 32 reviewers to assess ChatGPT's output. They were asked to determine if it is comprehensive, correct, and can contribute to the field of finance studies.

Read also: Fake ChatGPT Apps Start to Steal User Data: Delete These Apps Now

Considered Acceptable

Lucey and Dowling claim that all these studies were mostly considered acceptable by the panel. This could indicate that the AI tool may be capable of producing quality academic research ideas even though it is still in its infancy. 

They discovered that different research areas received varying ratings. The dataset and the research proposal were frequently given good ratings. The rating for the literature reviews and testing recommendations was lower but still acceptable.

The researchers suspect that ChatGPT is good at taking a set of external texts and connecting them, but it is also weak in complex stages, especially when there are a lot of stages involved in the conceptual process, such as literature reviews and testing.

"For now, we think that researchers should see ChatGPT as an aide, not a threat. It may particularly be an aide for groups of researchers who tend to lack the financial resources for traditional (human) research assistance," the researchers wrote in the article.

But they also note that academics must be aware of the ban on its use when writing journal papers. It must be used with care because there are varying views on this tool's implications for the academic world.

The findings of the team were published in the journal  Finance Research Letters.

Related Article: Google Invests in ChatGPT Competitor as It Joins AI Race, Gain Access to Claude

Byline

ⓒ 2024 TECHTIMES.com All rights reserved. Do not reproduce without permission.
Join the Discussion