BETTER IMAGES OF AI
In a world where searching for 'artificial intelligence' yields clichéd images of humanoid robots and glowing brains, the need for a more accurate visual representation is paramount. Recognizing this, AI4Media, in collaboration with Better Images of AI and AIxDesign, launched an artist commission Open Call to cultivate a nuanced visual language for AI. This initiative challenged artists to rethink the portrayal of AI, moving beyond entrenched stereotypes to more responsibly reflect its role in society.
Responding to this call, collaborator Anne Fehres and I were one of three selected artists to contribute a series of work to the ‘Better Images of Ai’ image library. Our collaboration resulted in four digital collage artworks, each addressing key aspects of AI's interaction with human life. The artworks, developed in collaboration with We and AI for Better Images of AI, can be explored at Better Images of AI. "Hidden Labour of Internet Browsing" uncovers the unseen human input in everyday digital interactions, such as reCAPTCHA verifications. "Humans Do The Heavy Data Lifting" spotlights the significant yet often invisible human labor in online content moderation. "Models Built From Fossils" metaphorically represents AI data as 'fossils,' drawing attention to the biases and contexts embedded within AI models. Finally, "Data is a Mirror of Us" contemplates AI's role as a societal reflector, capturing both the merits and flaws of our technological creations.
The aim of these images is that they can function as catalysts for a broader conversation about AI. Our vision for the project aims for these images to be utilized across diverse media platforms, fostering a more nuanced and effective understanding of AI. As AI continues to evolve and integrate into various aspects of society, our works aim to play a crucial role in shaping the public's perception, contributing to a more informed and balanced view of this rapidly developing field.

This image is exploring the hidden AI labour of everyday internet usage. An everyday example of this is the use of reCAPTCHA systems. Occasionally as we navigate the internet we are asked to prove we are not a robot, by being asked to identify a series of objects. Typically it might be crosswalks, traffic lights, bicycle riders or palm trees. Our small contribution cumulatively contributes to a large portion of data labelling, which is used in monetized products.

This image highlights the role of human labour, rather than AI, in maintaining ‘appropriate’ virtual experiences. Societal imagination of AI is fueled by fantasy more than reality - there is a deliberate strategy for when to and when not to reveal the human labour within AI systems. The role of the ‘content moderator’ is often overlooked and assumed to be an automated system. There is a blurring between human and technical achievement. Content moderators are doing a lot of ‘heavy lifting’.

The aim of this image is to highlight that data, in both its ‘raw’ and reinterpreted form in AI systems, has a human origin. While moving through different seemingly abstract systems, a recognition of this origin highlights the notion of “human in the loop” and pushes back against the notion of AI as a ‘black box’. This work explores this idea through the symbolism of the mirror and reflection.

This image is exploring the hidden AI labour of everyday internet usage. An everyday example of this is the use of reCAPTCHA systems. Occasionally as we navigate the internet we are asked to prove we are not a robot, by being asked to identify a series of objects. Typically it might be crosswalks, traffic lights, bicycle riders or palm trees. Our small contribution cumulatively contributes to a large portion of data labelling, which is used in monetized products.
This project was commissioned by AI4Media, AIxDesign and Better Images of AI funded by the EU’s Horizon 2020 program. Coordination of the project was led by a talented team including Program Lead Ploipailin Flinn and Creative Lead Nadia Piet.
The final works can be explored at Better Images of AI.