The Balance Between Human Capital and Artificial Intelligence: The Use of ChatGPT in Scientific Communication

Authors

DOI:

https://doi.org/10.5281/zenodo.14728109

Keywords:

digitalization, human skills and abilities, automation, cosine similarity, text coherence metric, data processing, adaptive algorithms, cognitive analysis

Abstract

The development of artificial intelligence (AI) can be considered another revolutionary process. Its expansion into all areas of human life is only a matter of time. Today, numerous industries, such as science, medicine, shipping, and business, already utilize AI daily. Furthermore, as demonstrated in this study, artificial intelligence is increasingly employed by students and researchers.

At the same time, alongside the evident positive practices and outcomes, such as process automation, accelerated search, and calculations, there are also negative aspects of AI usage. This study analyzes whether humans possess the necessary skills for the correct utilization of AI and whether it can be reliably applied in scientific communication, where accuracy and reasoning are of utmost importance.

Objective: To conduct an experiment and evaluate the ability of AI, particularly ChatGPT, to generate responses that meet the requirements of scientific texts. To assess the ability of AI to produce text that approximates scientific style and structure, while also reflecting the ideas, theses, and emphasis of the original author. Based on the results, the study aims to analyze the necessary skills and competencies required for the effective use of AI in scientific communication.

Methods: The study employed mathematical analysis methods, including vector computations, to determine the similarity between texts. Additionally, analysis, classification, and modeling were used to evaluate key textual characteristics for comparing the outputs generated by ChatGPT with the original text.

Results: The experiment provided visual data on the correspondence between the text generated by ChatGPT and the original text. The level of similarity between the texts was approximately 50%. However, the AI was unable to produce a higher-quality result, indicating limitations in its ability to fully convey scientific ideas.

Conclusions: Despite the progress and development of AI, including ChatGPT, it is essential to critically evaluate the results it generates. Within the scope of scientific communication, an increasing number of researchers and scientists are turning to AI for assistance. However, it is crucial for individuals to develop new skills and competencies for the proper use of AI. This study presents a list of essential competencies that enable effective interaction with ChatGPT and other AI technologies.

Published

2024-11-29

How to Cite

Sienko, O., & Gnatyuk, B. (2024). The Balance Between Human Capital and Artificial Intelligence: The Use of ChatGPT in Scientific Communication. Current Issues of Economic Sciences, (5). https://doi.org/10.5281/zenodo.14728109