Democratizing algorithmic news recommenders: how to materialize voice in a technologically saturated media ecosystem

Abstract

The deployment of various forms of AI, most notably of machine learning algorithms, radically transforms many domains of social life. In this paper we focus on the news industry, where different algorithms are used to customize news offerings to increasingly specific audience preferences. While this personalization of news enables media organizations to be more receptive to their audience, it can be questioned whether current deployments of algorithmic news recommenders (ANR) live up to their emancipatory promise. Like in various other domains, people have little knowledge of what personal data is used and how such algorithmic curation comes about, let alone that they have any concrete ways to influence these data-driven processes. Instead of going down the intricate avenue of trying to make ANR more transparent, we explore in this article ways to give people more influence over the information news recommendation algorithms provide by thinking about and enabling possibilities to express voice . After differentiating four ideal typical modalities of expressing voice (alternation, awareness, adjustment and obfuscation) which are illustrated with currently existing empirical examples, we present and argue for algorithmic recommender personae as a way for people to take more control over the algorithms that curate people's news provision. This article is part of a theme issue ‘Governing artificial intelligence: ethical, legal, and technical opportunities and challenges’.

Governing artificial intelligence: ethical, legal and technical opportunities and challenges

“…Furthermore, there is a need for more non-US led initiatives like the Europe-based AI4People 4 and the Council on Europe's Expert Committee on AI and Human Rights. 5 is important to have more Europe-led initiatives, we must also incorporate concerns from the Global South. Marda's article about India highlights why these voices are especially relevant [27].…”

On the Democratic Role of News Recommenders

“…With a more liberal recommender, it would be perfectly acceptable to speak of information that the heterogeneous citizenry "wants to know." Therefore, a well-designed, diverse recommender would also incorporate a certain element of flexibility, allowing citizens to customise the recommendations to better reflect their interests and preferences, even if not all users will make use of that opportunity, a decision that would be fine as long as it constituted an expression of their autonomy (Harambam et al 2018). In fact, preselected choices, particularly when they do not allow citizens to understand why they have received particular recommendations, or do not provide them with the means to influence the settings, are suspicious from a liberal theory point of view.…”

Reading News with a Purpose

“…The third layer supports epistemic goals that foster self-actualization. Few studies have tried to connect transparency and explanation with certain personal or societal values and goals [15,22,23,38]. This layer departs from the simple information-finding of previous layers by promoting discovery and exploration.…”

“…Yet, the task of opening the black box of recommender systems remains notoriously hard to achieve [8,28]. For the domain of online news, this might limit the readers' control over information diets [15]. Limited explainability can also Conference'17, July 2017, Washington, DC, USA 2020.…”