In contrast, Ford (2016) investigated how the robotization of society has resulted in machines rather than human beings gradually performing different professions, implying that people increasingly tend to be substituted and replaced by robots and intelligent software.

Cevera (2017) studied the effects of AI, big data, and the internet on journalism through robots or bots. Like Ford (2016), he observed an evolution, which is divided between those who decide to adapt to the changes and those who choose to work in the profession traditionally.

Based on the polysemy of the concept of “artificial intelligence” Hansen, Roca-Sales, Keegan & King (2017) give an account of the forum that took place in 2017 by the Tow Centre for Digital Journalism and the Brown Institute for Media Innovation, in which questions relevant to the present and future of journalism were reflected on.

On the other hand, Lindén (2017) and Salazar (2018) use the case study(s) to approach artificial journalistic intelligence, just as Karlsen & Stavelin do in their academic paper (2014) due to the insufficient scientific literature in this field of study.

On the one hand, Karlsen & Stavelin (2014), conduct in-depth expert interviews in the six major Norweigan newsrooms as a comparative case study. They compare traditional journalism to computational journalism and reveal that computational or software journalism skills and tools vary from those used in conventional journalism, although the values and objectives are similar. They also find little evidence to be able to affirm that computer journalism is more effective or frees information professionals from specific technical aspects of their work.

On the other hand, Lindén (2017) identifies that automated news is more effective and satisfactory for workers because it allows them to steer away from routine tasks that could generate human errors. Nevertheless, it brings about consequent job losses while creating new job opportunities in AI linked to computational thinking.

Salazar (2018) uses case studies of journalistic initiatives but combines them with expert interviews to identify the advanta-ges and disadvantages from both a professional and ethical standpoint. In this sense, he determines a reality with a double dimension in this new form of journalism: new opportunities for the future and collaboration between man and machine that calls for an occupational redefinition.

Another technique used to approach journalistic AI is through experiments. In this line Matsumoto, Nakayama, Harada & Kuniyoshi (2007) describe the process and system of the journalistic robot: “(1) autonomous exploration (2) news recor-ding and (3)generating articles” through labels, which enables images with words to be described, thus making a notable contribution to “artificial journalism”.

Clerwall (2014) not only uses the experiment but also a survey of participants who access different news items written by both journalists and software. The questions revolve around variables such as the perception of quality, credibility, and ob-jectivity of the news items which subjects have seen. It aims to analyze how the audience perceives the content generated by software as opposed to the content produced by journalists. In this sense, he concludes that software-generated content is associated with description and boredom, even though they are considered to be objective.