The study “Imagination, Algorithms and News: Developing AI Literacy for Journalism” by Mark Deuze from University of Amsterdam and Charlie Beckett from London School of Economics framed the dimensions of Artificial Intelligence (AI) and presented the components of AI literacy.
According to the authors, journalism is intimately connected to the role of AI in news industry. AI is an umbrella term that refers to a range of technologies such as statistical data analysis, machine learning, and natural language processing.
The benefit of AI is that it is extremely good at repetitive, simple tasks, but the downside is that it suffers in complex tasks as it is vulnerable to manipulation and mistakes and tends to be systematically biased due to amplifying existing inequalities. Thus, there are ethical concerns.
The authors categorize AI as being related to the workflow of journalism, being a potential topic for journalism, and related to the business of journalism. It is also said to be an amplifier of all existing issues, such as winning and losing trust, covering diversity and inclusion, and how it assumes responsibility.
The taxonomy that the authors use consists of three parts, based on which they offer their solutions. 1) the business of news industry tends to emphasize instrumental rather than imaginative approaches to AI, 2) industry and academic discourse problematizes the technology yet struggles to come to terms with the ethical implications, and 3) the historical fallacy of association perfection with machines, that blinds the viewers to the co-construction of humanity and technology.
AI literacy, then, is not just about knowing about AI, but also about the normative dimensions, and about knowing how to apply AI responsibly, creatively, and efficiently. Symmetrically, the authors also present three components for this, knowledge, ability, and skills:
- knowledge, including genealogy and avoiding fantastic and dystopic claims;
- ability to recognize instances where AI is useful and can be applied creatively, and when it should be avoided; and
- skills to teach and guide others in understanding, imagining, developing and implementing AI.
Both of the authors have a track record of working with the research of AI in journalism, so their suggestions will likely gain wider traction in this current and important topic.
The article “Imagination, Algorithms and News: Developing AI Literacy for Journalism” by Mark Deuze and Charlie Beckett is in Digital Journalism. (limited access).
Picture: Untitled by @deepmind