Affiliation:
1. Beijing Normal University, China
2. University College London (UCL), UK
Abstract
Artificial Intelligence is a term used frequently in academic and other writing, but do we have a clear understanding of what it means? This article starts from first principles, taking a dialectic approach, to raise questions rather than give prescriptive answers. It unpacks some specific examples of the use of AI in journalism and automated approaches to news reporting. The manipulation of media has become commonplace and of greater interest as information itself can be used as an effective weapon to sow confusion and disruption, socially as well as politically. AI depends on the training data and modelling, but the sampling and engineering is done by humans with all the potential for bias, whether intentional or not. Biased datasets and the potential for uncertainty are constant dangers; we need to understand both the data and the processes that go into the AI-driven results, and always be prepared to question everything.