Trust of journalists in collaborating with AI

Abstract

Literature reviews indicate that journalists use ChatGPT as an advanced search engine that provides ready-made answers to specific questions, generating text in a predetermined manner. Although the open version of ChatGPT does not provide sources of information, the way AI accomplishes its tasks can be inspiring for experienced journalists, aiding in the creation of valuable content. However, artificial intelligence (AI) cannot replace high-quality journalism or independently create valuable journalistic content. Moreover, AI lacks the capability to make moral decisions, which are often essential in journalistic work. Nonetheless, collaboration between journalists and AI can increase the likelihood of achieving the goals of quality journalism. This chapter presents the results of empirical research based on interviews with Polish journalists. The research reveals journalists’ limited trust in generative AI and outlines their beliefs regarding the ethical determinants of such collaboration. The limited trust stems from the fact that AI-generated sentences, often short and simple with a dispassionate exposition of details typical of informational genres, may form the basis of texts devoid of evaluation, opinion or commentary. However, these texts require verification by journalists, as AI can mislead through arbitrary and biased selection of facts. Analyzing the ethical determinants of journalists’ cooperation with AI can serve as a foundation for developing best practices in this area.

Ähnliche Arbeiten