Informative disorders in initial digital circulation news articles on ChatGPT (#423)

Authors

Alarcón Llontop, Luis Rolando

Lomas Chacón, Pablo

Cruz Páez, Pamela

Pasapera Ramírez, Sindy

Torres-Mirez, Karl

Abstract

The research aimed to quantitatively and qualitatively evaluate the occurrence of informative disorders in news articles of digital circulation on the ChatGPT artificial intelligence application. Under the integrationist paradigm, mixed approach, from an exploratory level and hermeneutic typology, 78 news pieces obtained in a month as Google alerts were analyzed, from which general aspects were probed, such as the journalistic genre and its origin, and specific aspects within the framework of the concept of information disorders. We resorted to the techniques of content and discourse analysis; and to a double instrument: observation/coding sheet. It was found that, although the percentage of initial information on ChatGPT does not fall into the categories framed within the possibilities of informational disorders, there are - first in number - some erroneous (10), others incorrect (06) and, finally, a handful of false (04). In conclusion, the voracious emergence of the application in its early stages has not been a pretext for general production oversights and misinformation about it, which is positive in principle, but the margin of the informative disorders found, even minimal, calls attention to the care that must always be taken in journalism, especially with topics of enormous interest, in a context in which the essential principle of journalistic truth today more than ever does not have to accept concessions.