Charismatic Megapigment video. Charismatic Megapigment on exhibition at EMERGENCE.
Charismatic megapigment is a collaborative project produced with Charley Peters and Wesley Goatley. In this installation, a green painting by Charley Peters is analysed in detail by a robotic camera and screen roaming across its surface. This camera feeds a machine learning system, built by Wesley Goatley, which is trained on a custom-made database of 100,000 images classified as ‘green’ by Google.

As the camera slowly moves across the painting it draws and displays a nearest-neighbor match to the part of the painting it is ‘seeing’ to the bank of green images from Google. In doing so it creates the impression of investigating or interrogating the painting, looking for meaning in the abstract forms and colours. However, the painting and the images are contextually unconnected and any meaning or inference drawn from the computer’s analysis of the painting is purely an inference of the viewer. 

The project critiques an aesthetic meaning-making tendency in both machine-learning and greenwashing. Machines’ ability at performing pattern-recognition is often held up as s sign of some sort of intelligence or meaning-making quality. Equally, contextless images of fields, forests and smiling children are often intepreted as ecological-consciousness. The juxtaposition of these two ways of deploying visual culture to aid a technological and political idea is exposed as critically unsound through the project. The project also seeks to technically explore the limitations of pattern recognition and meaning-making in machine learning. The generative nature of the outcomes and the sendipitous matches of Google images and painting invite the audience to consider the hand of the artist, developer or engineer in making editorial and critical decisions about what consitutes a ‘match.’ 

Further reading:
Hayles, N. K., (2019). Can Computers Create Meanings? A Cyber/Bio/Semiotic Perspective. Critical Inquiry. 46(1), pp. 32-55.
Pohflepp, S. (2016). Pattern Agnosia and the Image not made by Human Hand.
Steyerl, H., (2018). A Sea of Data: Pattern Recognition and Corporate Animism (Forked Version). In. Bachman, G., Beyes, T., Bunz, M., & Chun, W. H. K., (Eds.) Pattern Discrimination. Minneapolis: University of Minnesota Pres.

09.2019 – EMERGENCE, London Design Festival.


Last updated: 05.2020