Abstract
News automation and the path towards hybrid journalism
Right at the beginning of the book, Nicholas Diakopoulos sets the tone for what comes next: algorithms, datafication, artificial intelligence and automation – it is all about journalism becoming increasingly hybrid and how to design such a journalism in the future. He warns as follows: ‘Designing hybrid journalism won’t be easy; there’s no cookbook, no algorithm here. Yet production processes will need to be reinvented to take full advantage of technical capabilities’ (Diakopoulos, 2019: 35). Diakopoulos continues to whistle the tune of hybridity throughout the whole book by focusing not only on ‘who’ actually does journalism but also on ‘what’ does journalism. This means adopting a perspective that takes into account both human actors and nonhuman actants and to analyse how the two engage; how they are assembled in newswork; and how processes, practices and workflows are iteratively and creatively re-engineered (by humans). In the area of automation and artificial intelligence particularly, where technological determinism and ‘boosterism’ is just around the corner, Diakopoulos adopts a pleasingly dispassionate but nevertheless optimistic stance, acknowledging that (human) journalists in the future will – and must – have a lot to add when it comes to crucial aspects of their profession such as expert knowledge, complex communication and, not the least, ethical judgement.
The book’s chapters concentrate on six different contexts where automation – and its big brother, datafication – plays a crucial role. After laying out his thoughts on the design of hybrid journalism in the first chapter, Diakopoulos covers topics such as data mining (Chapter 2), automated writing systems (Chapter 3), newsbots (Chapter 4) and distribution algorithms (Chapter 5). In all these areas, and with an exceptional number of examples and small case studies, the author demonstrates how technology and automation are able to help and supplement journalists in the newsroom, without replacing them – showing also the limits of implementing new ground-breaking technology in complex social settings such as newsrooms. In Chapter 6, the author concentrates on the many issues of automation and algorithms by looking at a specific journalistic beat: algorithmic accountability reporting – ‘[…] a reorientation of the traditional watchdog function of journalism toward the power wielded by algorithms’ (Diakopoulos, 2019: 207). This beat becomes crucial if one wants to unblackbox the immense power algorithms have already gained in society.
One area where automation in journalism and the datafication of society go hand in hand is data mining. This technique is not only used to comb through massive amounts of data in order to identify stories of editorial value. Data mining algorithms can also be used for predictions, to detect and monitor events, for verification purposes and fact-checking, and for content curation. The author describes both the opportunities of the technology as well as its issues. In the case of data mining particularly, many ethical issues can arise when it comes to the way large stores of data are used. Journalists may have to justify their editorial output, and they may be held to account for the use of personal and sensitive data. However, algorithmic governance and accountability in news media are rarely implemented and incentivised (Diakopoulos and Koliska, 2017; Porlezza, 2018, 2019) Thus, one of the most fascinating parts of the book is the author’s discussion of algorithmic accountability in Chapter 6. He addresses not only how the journalistic discourse surrounding the technology evolves as a beat, but also what kind of approaches can be implemented for an ethical and accountable use of algorithms in the news media by specifically referring to the concept of transparency. All too often, transparency is seen as an all-or-nothing binary concept, but Diakopoulos manages to offer an insightful analysis that goes beyond the simplistic request for more transparency of algorithms, demonstrating that, particularly in journalism, it has to be weighed against other professional norms.
Albeit still in a nascent state, automated content creation is perhaps the most common thought when it comes to the automation of news. This is not limited to the generation of texts, but the inputs can also include visual content such as photos and videos. In fact, this is not only the context where the roles and tasks of human journalists may shift the most while accommodating the technology’s design, but also the area where the audience has to be ‘willing to accept automated journalism as a legitimate mode of cultural production’ (Carlson, 2017: 155). This becomes even more pressing when it comes to the interaction with news or chatbots, a tool news media organisations are increasingly experimenting with (Ford and Hutchinson, 2019), and brings ‘forward a series of issues and effects upon journalistic narrative, content and professional practices’ (Veglis and Maniou, 2019: 4). Eventually, journalists will have to tackle the question what these technologies mean to them and what kind of consequences the automation of news will entail for their legitimacy and authority.
Overall, the book offers a well-researched and comprehensive statement about the current (quickly and incessantly evolving) state of news automation and its implications for newswork as well as the journalistic profession. As the technology’s capabilities continue to grow at high speed, the main question journalists and news organisations will be confronted with is how humans and non-humans can and will co-evolve. In other words, how can newsrooms design hybrid journalism in the future, where both technological actants and journalistic actors complement each other, and where they mutually shape and manage workflows and media innovations (Westlund and Lewis, 2014)? This is what news organisations have to aim for, because it ‘will boost the productivity, quality, and scope of journalistic efforts, while ensuring human autonomy and agency in what is ultimately produced’ (Diakopoulos, 2019: 245). A must-read for all those scholars interested in news automation, datafication and the algorithmic accountability of journalism.