What a Machine

What a Machine is part of the Attention Machines (AM) Project which aims to test how works of fine art elicit attention differently from more mainstream digital media culture. Recent commentary on the tech-sector alleges that tech-corporations mine users’ attentional capacities by harmful habitual user-experience (UX) design. Social media UX has been attributed to attentional deficit in users, leading to mental health risks, such as anxiety as well as social polarization. The AM project aims to find ways to leverage formal elements in UX to produce engagements with viewers / users that enhances rather than diminishes viewers’ attentional capacities.

What a Machine is a solo exhibition of paintings and creative code that incorporates Artificial Intelligence algorithm structures as a method for iterating new generations of AMs in order to optimize their capacity for helping to build attentional resilience. Building on the AMs generated in the author’s previous exhibitions, the works in this show introduce Chromatic Attention Machines (ChAMs), which use a wider colour gamut. The author developed five generations of ChAMs using the operations found in the Genetic Algorithm (Holland, 1992). Rather than using a computer based-AI engine, the author performed operations from a GA in the studio, mimicking generator, selector, fitness, recombinator and mutator functions to increase the attention-grabbing and attention-holding qualities of the works. The resulting works demonstrate the efficacy for the GA as a structure for artistic iteration in impact-driven art, by maturing the material manufacture, and discrimination functions in line with the studio mission.

What a Machine was publicly exhibited at Griffith Regional Art Gallery in NSW. The exhibition contributed to public awareness of the potential for AI for studio artists navigating social issues in their work. This public awareness is evident through regional media interviews with commercial radio at Triple M and Southern Cross News. The works exhibited in What a Machine also resulted in a sample size sufficient to empirically test the efficacy of AMs in experimental conditions at UTAS.