Nazaj v prihodnost: avtomatizacija in preobrazba novinarske epistemologije

Abstract

The article stems from the historical discussions on subjugating journalism to the capitalist necessity of "rationalisation" of the labour process through technological innovations. As the continuation of these tendencies, the gradual implementation of algorithms to solve journalistic problems based on data, thereby supplements, extends, or replaces journalistic work with technology. The apotheosis of this process is automated journalism, which, with or without human involvement beyond the development of algorithms, asserts itself in the production of news that is generated and adapted on the basis of always updated highly structured databases. Although news automation primarily complements the established labour process, the historical tendencies of technological innovations in the newsroom, the continuous development of natural language generation systems, and the technological drama surrounding automation raise challenging questions about what and how we know via journalism and how that knowledge is substantiated. In automated journalism, the construction of "objective" facts and its language game are conditioned not only by the combination of technology and human activity, but also by the separation of technological objectivity and human subjectivity. The automated process of datafication of facticity of news, in fact, draws from the authority of algorithmic judgment, which is concentrated in the aura of objectivity of data and the principle of legitimation through instantaneous, direct and authoritative performance. In this respect, the historical principles of "objective" journalistic work are subjected to qualitative change. While technological innovations throughout history have caused that journalists have gradually lost control of their own work, automation potentially eradicates journalistic judgment that is institutionalized in the patterns, attitudes, and practices of the modern newsroom.

Habitual Generation of Filter Bubbles: Why is Algorithmic Personalisation Problematic for the Democratic Public Sphere?

“…Recommending algorithms were often treated as a counterpart to human activity, especially in the field of journalism and media studies, since they perform the selection of content that is delivered to users and consequently replace the selective function that was traditionally performed by humans, professional editors and journalists in particular (Carlson 2018;Nechushtai and Lewis 2019;Vobič, Robnik-Šikonja, and Golob 2019). Algorithmic personalisation thus "challenges the continuing relevance of established theories of journalistic gate-keeping" (Thurman 2011, 395).…”

Hume’s Empiricism versus Kant’s Critical Philosophy (in the Times of Artificial Intelligence and the Attention Economy)

“…"AI-powered recommendation systems" have become an increasingly key focus of interest as AI has "recently reached the level of development that makes their functioning comparable to human thinking and allows them to perform tasks requiring (close to) human intelligence" (Sorbán 2021). Such assumptions, however, have not only remained at an abstract level, but have to a certain extent also materialised in a very concrete way in the modern world, as algorithmic content selection is now replacing some of the professions that are closest to the ideals of human curation and decision-making, such as journalists and editors (Carlson 2018;Vobič, Šikonja, and Kalin Golob 2019). Artificial intelligence "which is getting more and more independent from humans through machine learning methods", could, in this context, be considered "a major threat to the human race" as seen in "countless science fiction movies in which AI machines try to take control over mankind" (Hosseinpour 2020, 49-50).…”