- - - - - - - - - - - -

Title: tbc by algorithm

Artists: ██████ █████ ████████ ████

Curator: Jon Smeathers

Group research: December 2023 – March 2024

Exhibition: 13 April – 18 May 2024

“All models are wrong, but some are useful”

- George Box

A speculative project centralised around the generation of a curatorial algorithm.

The ubiquitous integration of algorithmic structures within contemporary society has given rise

to an emergent phenomenon known as algorithmic culture. This cultural paradigm is changing

the contours of society through its control logics and statistical governance that is resulting in a

re-synthesis of human oriented sociocultural attributes. In the realm of creative practices, this

entrenchment of algorithmic structures has introduced a transformative shift, fundamentally

altering the landscape of aesthetic discernment, artistic curation and causing an amplification of

data extractivism. The ability of an algorithm to undertake intensive cyclical learning

approaches, to analyse evolving variations of data sets at light speed alongside the

computational indifference in binary problem-solving is imposing a technological decisionism

while also becoming a conduit for thinking through curatorial methodologies that escapes previous

human orientated limitations.

The algorithmic paradigmatic shift in curatorial methods fosters a departure from traditional,

taste-driven methodologies that often perpetuate elitist conceptions of artistic value or contain vast

knowledge gaps. Not without its systemic problems, algorithmic curation enables guiding

principles to transcend conventional aesthetic canons and groupings, subvert decision making

processes, an ability to alter ingrained bias’ and has the potential to show us something outside

an individual humanalogue perspective. Although, does an algorithmic curation operate with

care or simply does it value making a clear decision quickly more so than it does a thoughtful

one?

Four artists from across Australia have been invited to become ghost workers* in a research group

that will critically analyse the sociocultural implications of algorithmic structures and build a model

for exhibition making. Discussion topics will include ethics, labour considerations, potential biases,

and techniques for data reduction within algorithmic curation. Based on the collective insights

garnered from these discussions, the algorithm will learn to embody and emulate the embedded

categorisation processes of the four individuals alongside the collective’s perceptual boundaries and

their modes of engagement within gallery-based sociotechnology infrastructure.

The algorithm’s dataset will be drawn from interactions across the project’s lifespan including, but

not limited to, private meetings, internal documents, image drops and a public chat forum. The

resulting algorithm will: produce a complex model for considering each artists mode of engagement

within specific exhibition conditions; offer an apparatus for an audience to see how an algorithm can

distinguish the unique qualities of an artist; and, produce a generalised curatorial framework for

future exhibitions that will materialise participants cognitive bias and decision-making processes

even after their death.

A chat group will be utilised by the artists/ ghost workers for testing out and constructing datasets.

The chat group will be updated on an ad-hoc basis. All data entered will be embedded into the

algorithmic dataset. The impact will vary depending on the weighting / filtration systems developed

throughout the project.

*A ghost worker is a human who performs short-term tasks on demand, anonymously, through

automated platforms and the work is disguised as being automated. Artist names have been

redacted to keep in line with the anonymity of a ghost worker.

Project is supported by Contemporary Arts Tasmania: Project Mentor