214 | 29, pp. 213-233 | doxa.comunicación

July-December of 2019

Algorithms and bots applied to journalism. The case of Narrativa Inteligencia Artificial: structure...

ISSN: 1696-019X / e-ISSN: 2386-3978

1. Introducción

Artificial Intelligence (henceforth AI) occupies a place in newsrooms dominated by people (Túñez; Toural; Cacheiro, 2018; Vállez; Codina, 2018), as it generates greater productivity and efficiency in content production (Papadimitriou, 2016). It involves computer systems fed by journalists using data and algorithms to automatically generate comprehensible news for the audience from a previously programmed structure and formula (Barrat, 2013; Bunz, 2010). Consequently, its use has affected the routines of news professionals, mechanizing their search, classification and information processing functions (Lemelshtrich, 2018; Lindén, 2017). This trend is already taking place in Anglo-Saxon countries, where it is called robojournalism (Burrell, 2016; Carlson, 2015; Fernández, 2018; Lemelshtrich, 2018; Lee; Kim, 1998; Matsumoto et al, 2007; Van Dalen, 2012), although Bercovici (2010) prefers to call it “algorithmic journalism”; Clerwall (2014), Karlsen and Stavelin (2014) speak of “computational journalism”, and Caswell and Dörr (2018) talk about “automated journalism”. In Spain, Salaverría (2014) has named it “high-tech journalism”. The origins of AI go back to the advances made by Alan Turing in message decoding during World War II (Rusell & Norvig, 2003). The term as such was first used in 1950, but it was not until the 1980s that research began to grow by solving algebraic equations and analysing text in different languages (Sancho Caparrini, 2018; Van der Kaa and Krahmer, 2014). Podolny (2015) associates its decisive expansion with the increasing demand for news from the information society in the digital age, which insists on content with an appetite that humans can no longer satisfy. Coddington (2015) and Van Dalen (2012) explain that the journalistic profession is becoming ever more deeply involved in the trend toward commercialization, and is being more heavily based on business logic. Therefore, the automation of tasks in journalism is adjusted to the predisposition to obtain higher profit margins and lower production costs.

However, it was not until the global economic recovery took hold in recent years that investment in this field has increased (Salazar, 2018). As a result, more and more media are betting on this emerging technology that handles big data and draws conclusions with extreme speed and precision in its results (Kim et al., 2007).In any case, this trend will continue in the coming years as pointed out in the latest report by the Reuters Institute entitled, Journalism, Media, and Technology Trends and Predictions (Newman, 2018). Eudes (2014) has written that this is the beginning of a great adventure, as automated

using in-depth, semi-structured interviews with four company professionals with complementary profiles, as well as participant observation during the process of creating an automated news item, and an online questionnaire filled out by 145 journalists (N=145) to assess the quality of the texts produced. The results show that the start-up company fulfils the demands of different sectors such as communication media, and has been increasing the productivity of its content, although it still does not generate complex texts, as this is one of many human qualities that robots continue to lack.


Artificial intelligence, Algorithmic journalism, Automated journalism, Production, Informative quality.

rrollado a lo largo de seis meses, mediante entrevistas semiestruc-turadas en profundidad a cuatro profesionales de la compañía con perfiles complementarios, observación participante durante el proceso de creación de una noticia automatizada y un cuestionario online cumplimentado por 145 periodistas (N=145) para valorar la calidad de los textos que produce. Los resultados muestran que la startup cubre las demandas de diferentes sectores, como los medios de comunicación, incrementando la productividad de contenidos, aunque todavía no genera textos complejos, una cualidad humana de entre las muchas de las que adolecen aún los robots.

Palabras clave:

Inteligencia Artificial, Periodismo algorítmico, Periodismo automatizado, Producción, Calidad informativa.