Current AI Visuals are one a one-size fits all sci-fi fantasy, highlights OII doctoral researcher Maggie Mustaklem. In her latest blog she explains why as AI’s remit grows, this needs to change.
AI is not new. In fact, “narrow AI” has been with us for quite some time, quietly integrating itself into our algorithmic feeds, GPS and Alexas. The current AI gold rush, with Generative AI and Large Language Models (LLMs), is not happening quietly, weaving its way into our lives with much fanfare.
Right now, media coverage about AI is everywhere. As AI moves beyond applications within Big Tech and into industries ranging from healthcare to manufacturing, there is an enormous amount of content about AI. It’s pervasive in the news and on social media.
Yet despite incredible advancements in the technology, and incredible efforts to publicise its increased role in our lives, there has been woefully little to advance visual communication and content about AI. As AI technologies evolve, the imagery used to communicate media content about AI remains stubbornly static. It is persistently presented as blue brains and robots, often against a backdrop of blue code. Why is AI always blue?
Better Images of AI and Neema Iyer brilliantly identified tired AI tropes. If you run a story about generative AI, there are no robots involved. Why would your hero image include a pensive robot? If you ran a story about apples you wouldn’t include a pear photo as your hero image – it would confuse your readers. These inaccurate, masculine sci-fi fantasies are misleading. AI is not gendered, and it is no longer a sci-fi fantasy. It’s here. We use it every day. Despite its prevalence, AI remains an opaque, confusing technology for most of the public. Improving visual communication about AI would provide much needed support to improve public understandings of the technology.
Some organisations are tackling this problem head on. Better Images of AI is building an incredible creative commons stock library of AI images created by artists. Deepmind worked on a project with artists on visualising AI. Beyond this great work, there is a dire need for broad scale, commercialised visual communication of AI. This area needs more attention and more funding to develop more descriptive, understandable visual communication about AI. As AI moves into industries like healthcare, there is also a need to support industry specific visuals.
In addition to considering industry contexts, brands need creative content that supports precision around the various AI technologies they are implementing. Computer vision? LLMs? NLP? Machine Learning? Visuals and content need to align with the AI technologies brands are actually implementing. They also need to support the case specific uses of AI. Is it supporting an administrative function? Is it consumer facing? In order to appropriately situate AI’s industry placement and use case, AI researchers should be included in the development of visual communication. As the AI landscape continues to grow, hopefully we will shed a one size fits all approach to AI’s visual communication.
Find out more about doctoral researcher Maggie Mustaklem’s research which focuses on the implications of AI in design.
