Not Allowed for Algorithmic Audiences, 2021
4K CGI VIDEO ● 3D MODEL ● VOICE OVER ● PRINTS ON ALUMINIUM ● VINYL LETTERS ● SOUNDSCAPE ● ORIGINAL TEXT
Edition of 3 + 2AP.
video + portfolio request to email@example.com
Internet users upload tons of content on a daily basis. Around 3.7m new videos are uploaded every day on YouTube only. In these videos sounds produced by the human body are collected, categorized and labelled to build extensive audio data sets used to train algorithms. Voice user interfaces employ artificial intelligence to interact with users and execute commands. Virtual assistants, such as Siri, Google Assistant, and Alexa, are examples of VUIs.
'Not allowed for algorithmic audiences' is a video installation featuring also a series of tangible elements interconnected to the narrative. In the video a fictional virtual assistant under the wake word name ‘Voice’ exhibits an odd behaviour. They borrow an avatar and appear in front of their users. The virtual assistant goes into seven monologues for seven consecutive days in a hot August, before their patent expires and they get kicked out from the Athenian apartment to an e-waste dump. While active, Voice has managed to scan the entire contents of the Internet and gather all sorts of information, information that they long to share. The virtual assistant uses their seven brief monologues as an opportunity to introduce themselves. They talk about their skills, their ancestors, the rare earth elements they are made of, as well as voice and its significance. They reveal information regarding the listening infrastructures as well as the social dysfunctions and stereotypes, on which their programming and operation are based. Listening infrastructures, surveillance and the climate crisis are also part of the narration. Just before they reach the end of their monologues, in a final effort to reconcile humans and machines, they share tips with humans on how they can manage… not to be heard by algorithms online.
The feminist philosopher Rosi Braidotti describes voice as “a unique audio footprint of a human’s soul”. Speaking is an integral part of everyone’s individual identity, it carries the past and present. In the last years the usage of voice UIs is on the rise. The number of users for voice assistants, especially smartphones and smart speakers, is increasing among all age groups. What are these voice assistants to us? Buddies, servants or supervisors spying on voice’s biometrics? How are they trained to detect our physical and emotional states in order to predict our desires, fears, sickness and needs, harvesting at the same time our behavioural surplus? How do voice UIs and listening infrastructures reflect the past in the gender and race bias that they carry in them?
In her this work Kyriaki Goni integrates in a fictional narrative the current wealth of research on artificial intelligence and voice interfaces, as well as the relationship between humans and machines. Could a playful and mutually beneficial relationship between human and machine exist? Could this relationship be reinvented not as a relationship between product and consumer, but rather as two distinct and non-hierarchically ordered species. What would this post-anthropocentric approach mean for the power structures in our societies?
The virtual assistant’s avatar is modelled after the face of the Greek actress Sofia Kokkali, who also lends them her voice.
THE BREEDER SOLO SHOW
WORKS IN FOCUS, GENERAZIONE CRITICA
WENN DIE SPRACHASSISTENTIN WARNT, TAGESSPIEGEL
如何抵抗監控資本主義：智慧助理的獨白, VOICES OF PHOTOGRAPHY
THE IMPORTANCE OF HAVING A VOICE IN THE AGE OF SONIC SURVEILLANCE, FEROCIUS URBANITES
THIS ARTWORK WAS DEVELOPED WITHIN THE FRAMEWORK OF THE ARS ELECTRONICA ARTSCIENCE RESIDENCY ENABLED BY ART COLLECTION DEUTSCHE TELEKOM IN PARTNERSHIP WITH JOHANNES KEPLER UNIVERSITY.
THE WORK IS PART OF THE ART COLLECTION TELEKOM.