The current wave of pervasive automation and the ever faster development of technologies of prediction are radically transforming and reorganising all forms of labour. As the delegation of tasks to machines becomes widespread, Sanela Jahić’s starting position in this project was to explore the possibility of a machine-conceived exhibition. Can her labour as an artist be automated? The fact that an algorithm is created to interpret the individual’s relation to technology, and the way the latter influences the nature of labour, clearly puts the machine in a particular loop.
CHAPTER I: The Labour of Making Labour Disappear
Furthermore, as algorithms ripple through and integrate deeper with our work and daily lives, future projections on work note creativity, flexibility and innovative thinking as key anticipated skill demands that will be required of employees. These skills, which have traditionally also been attributed to the profession of the artist, are supposed to be among those most resistant to or the least exposed to automation. Many people think that what they do requires creativity, and that can never be expressed in the form of executable code or emulated by a machine. Contemporary perception usually sees creativity as something new, divergent and original, something that takes people by surprise. And yet, many tasks which might entail human faculties such as intuition, empathy and creativity, are already being outsourced to automated systems that simply perform them differently.
There are many ways to distil something into data points. Jahić converted her labour as an artist – her works, research and interests of the past 16 years – into data. As data, the features of her artworks become tables of numbers; each creative decision emerges in a row of digits. She then turns the decision making over to a predictive algorithm. The machine uses the dataset to sift through and identify patterns in her artistic labour in order to predict the content and aesthetics of her next artwork.
For the development and programming of the predictive model, Jahić worked with Iztok Lebar Bajec and Jure Demšar from the Faculty of Computer and Information Science in Ljubljana. Formulated according to her artworks and her working methods, this particular technology of prediction will become more advanced and complex the more data it receives. It might then produce something surprising, something that surpasses our expectations and could be perceived or mistaken for the machine’s creativity.
The project’s first stage The Labour of Making Labour Disappear was exhibited at Aksioma Project Space in 2018. The exhibition included the algorithm’s dataset and prediction, as well as the first new artwork produced according to it.
Digital Punch Card
Digital Punch Card is based on the results of the predictive algorithm and shows a small part of the invisible and dispersed network of the online global workforce performing tasks on the crowd-based, micro-labour platform Microworkers. The users were asked to download and install a custom-built app, which tracked the timestamp of their work and the number of mouse clicks and key presses they performed on the platform during a 24-hour period.
The fragmentation of jobs into outsourced tasks and deconstruction of wages into micropayments established one of the most unregulated labour marketplaces and are directly linked to the principles of scientific management.
“Perhaps the most prominent single element in modern scientific management is the task idea.” (Taylor: The principle of scientific management, 1911)
“Crowdsource your micro jobs to more than 1,261,920 workers worldwide, completing 37,870,348 tasks.” (Web-based platform Microworkers, 2018)
From the obtained data on digital labour performed on the platform and taking into account the formal parameters identified by the algorithm as essential for the work, Jahić made two data-driven line drawings and two animations, which show the traces of work of the invisible, flexible and on-demand online workforce.
CHAPTER II: The Labour of Making Labour Disappear, extended
In the second chapter of the project, a series of photographs was additionally collected to expand the artwork Digital Punch Card. Microworkers took photos of their workplaces, tagged them with the name of the object in the photo that is the most personal to them and then uploaded the photos to the site the artist created. 190 submitted photographs unveil the working environments of people who perform tasks posted on the platform.
Each photograph is labelled with the name of the worker, their age, city and country, how much they have earned on the platform in total, how many microtasks they have performed and how much they have been paid per task on average. The extended version of the project was exhibited in the Ivan Grohar Gallery in 2019.
CHAPTER III: Uncertainty-in-the-Loop
Uncertainty-in-the-Loop, the final stage of this two-year-long project, builds upon The Labour of Making Labour Disappear. The piece explores the possibility of a machine-generated exhibition by turning an artistic career into data sets. In its initial stage, the project shows that the algorithm’s predictive power is limited in its forecast of future outcome because it presents a selection among ready-made choices. And even though this might be an original selection, as a calculating predictor of the future, it is still effectively weighed down in the past. Every step of the way, the capability to create new choices is heavily constrained by choices that came before. In other words, conceiving artworks from the same conceptual depository is not really imagining something new, but a narrow alteration of existing inputs.
To break away from the machine algorithm rolling out combinations of prior data and to avoid sparking a feedback loop, the artist – in the project’s final stage –
provided the machine a look at her contemporary investigations as an early window into the present disorganization of her thoughts.
More specifically, her work and research activity over a period of six months was tracked. Tracking and logging was semi-automated. A custom-made algorithm sequence, which was developed additionally, sorted and analysed approximately six months of real time data. Every word from every text that passed in front of her eyes expanded the database, opening new horizons of thought.
Technical, sociological and descriptive terms of this present rendition – a state where the artist has made no active decisions, really – are ranked by the frequency of their occurrences and how they interconnect with the existing parameters from the past over multiple layers. The algorithm then determines tomorrow’s artwork based on observations today with the past flickering in the rear-view mirror.
Pataka
On the basis of the predictive model, Pataka was created, essentially a work that listens to what our voices tell machines about us. The artwork is the result of an analytical approach to work – using the output of an algorithm to determine what the artist should do next.
Proceeding from the parameters that were underlined with high confidence levels and high values as being crucial for her next artwork, the artist started researching how the state of one’s mental health leaves a fingerprint on speech production and the voice. She consequently met and worked with Dr Jude Dineley, a postdoctoral researcher at the Chair of Embedded Intelligence for Health Care and Wellbeing at the University of Augsburg in Germany, and Dr Nick Cummins from the Department of Biostatistics & Health Informatics, Institute of Psychiatry, Psychology & Neuroscience, King’s College London, in preparing a piece about machines listening to our voices, with a particular focus on depression. Pataka uses video and dynamic data visualisations to inform the viewer about the effects of depression in speech and how artificial intelligence is being used to detect them.
The three-channel video shows four African grey parrots that were taught the trisyllabic sequence Pa-ta-ka, which is commonly used to test diadochokinetic rate (DDK). DDK is an assessment tool used by speech-language pathologists that measures how quickly an individual can accurately produce a series of rapid, alternating sounds. The tool emphasizes changes related to a lack of muscle control in speech affected by depression.
The visitor is able to experience the cacophony of the parrots’ sounds and variances in their speech as they seemingly interact with and play off each other. Parrots are known for imitating the voices of the persons teaching them to speak, thus Jahić decided to use the African grey parrots as proxies. One can learn more about the particular vocal exercise, vocal biomarkers, machine learning, speech patterns and mental health, as well as the differences between academic and commercial research, by scrolling through data animations presented on two accompanying displays, which are installed with custom-made tripods in the exhibition space.
CREDITS
Technical support: Andrej Primožič
Graphic design: Vasja Cenčič
Development and programming of the predictive model and data visualisation: Iztok Lebar Bajec
Development and programming of the predictive model: Jure Demšar
Programming: Umer Muhammad
Data analysis and animation: Jernej Lunder
Research insight and assistance: Nicholas Cummins, Jude Dineley
Dynamic data visualisation design: Peter Primožič
Assistance in filming and picture post-production: Toni Mlakar
Sound post-production: Julij Zornik
Production: Aksioma – Institute for Contemporary Art, Ljubljana, 2018-2020
Co-production: Drugo more, Rijeka
Partners: Škofja Loka Museum; Faculty of Computer and Information Science, University of Ljubljana
PRESS
Hana Sirovica: Rad u nestanku (Kulturpunkt, Kulturoskop, 25.02.2020)
RADAR-CNS researchers collaborate with artist on audio-visual exhibition piece
(RADAR-CNS, Newsroom, 07.10.2020)
Sara Blažić: Otvorena izložba „Nesigurnost u petlji“ slovenske umjetnice Sanele Jahić (Moja Rijeka, 06.11.2020)
Breathy, tense, tight, flat: a person’s voice can hint at depression, and AI can pick it up (Innovative Medicines Initiative (IMI), Newsroom, 12.11.2020)
Mojca Kumerdej: Nasedamo spektaklu vse bolj zmogljivih nerazmišljujočih sistemov (Delo, Sobotna priloga, 21.11.2020)
Aude Launay: Sanela Jahić, The Statisticized Artist (Furtherfield, 24.12.2020)
Realized in the framework of the Dopolavoro flagship of the Rijeka 2020 – European Capital of Culture project, with support from the City of Rijeka – Department of Culture, the Ministry of Culture of the Republic of Croatia, the Ministry of Culture of the Republic of Slovenia and the Municipality of Ljubljana
Photos: Domen Pal / Aksioma, Janez Pelko, Tanja Kanazir / Drugo more