THE SPECIFICITY OF PERCEPTION OF A TEXT WRITTEN BY A HUMAN AND A NEURAL NETWORK

Abstract

The article is devoted to the study of the peculiarities of readers' perception of journalistic texts published in a business newspaper and generated by a neural network. The research was based on a survey of 89 participants of different age categories. Respondents were required to read 2 news on the same topic and 2 articles on the same topic and determine who wrote each text: a person or neural network, as well as evaluate the texts according to five proposed criteria: information content, reliability, logic, literacy and readability. As a result, we came to the conclusion that using a neural network is not usable to write a high-quality text for publication in business media.