Story-ing AI – mini-narrative patterns of contemporary online science journalism

This study explores the range of discursive patterns used to present artificial intelligence as a revolutionary but controversial technology in online science journalism. It uses a triangulated dataset of over a hundred recent mini-narratives sourced from New Scientist, Nature daily briefings, and Scientific American to reconstruct typical storylines in the thematic domains of research, business, and society, and to map their narrative trajectories (utopian, dystopian). The qualitative analysis uses the categories of agency, sentiment, point of view, and news value to capture these outlets’ contributions to the evolving sociotechnical imaginary of AI technologies. While acknowledging some risks of AI technologies, elite commercial science journalism highlights the benefits and celebrates the scientific advancements produced with or by AI. Also, AI technologies are communicated strategically to increase newsworthiness, through diverse complications in storylines with oscillating sentiments and a focus on impacts and novelty. This tends to prime news recipients to accept the inevitable technological progress and normalizes algorithms as increasingly independent research-performing agents.