Tracking the Evolution of the AI Art Debate on Twitter

The third presentation in this hot IAMCR 2023 session is Frederic Guerrero-Solé, whose focus is on the discussion of generative visual AI (e.g. Dall-E, Midjourney, Stable Diffusion) and its relationship to artists. There is a veritable media panic, but also a genuine discussion about how such generative AI tools are drawing on existing, copyrighted art for their creations.

Artists have reacted to this spreading of digital creation with some degree of concern: it challenges their role in society, their opportunity to earn a living from their art, and their authorship of creative works; many artists see the use of their works as input to generative AI as an unethical violation, and conversely some of the more extremist actors on the AI side of the debate are actively using artists’ works in order to hurt them.

So what are the public discourses behind this question: who are the most influential users in these debates, on Twitter, and what arguments are they making for or (in this case, especially) against generative AI? The present project gathered tweets from a large number of anti-AI art hashtags, analysed the content of these tweets, and used the retweet overlap network approach to understand the structure of the discussion network.

The first finding from this analysis is that the number of tweets supporting AI art has grown, while those attacking it have declined, over time. There is also a growing focus on this debate for many users involved in it, possibly indicating a growth in polarisation within the debate. The social networks discussing this issue serve as a social laboratory tracing the evolution of public attitudes towards this issue, and capture the emergence of a Luddite-like response to a new technology (but this is not meant to be imply a value judgment on the debate).