2020 Issue No. 34 – Inhuman Algorithms

Algorithms are integral to a digital, networked, automated society. Thrown into the public spotlight by a certain high profile search engine, algorithms are increasingly recognised to exercise agency in practices such as governance, surveillance, online personalisation, medicine, design, high frequency trading, credit scoring and plagiarism. Computational machines make decisions about things, people, places and experiences, and humans learn to address algorithms.

Algorithms have inhuman capacities. They do not become distracted, tired, impatient or emotional. At the same time the algorithm’s inhuman abilities can be understood as a desirable improvement on human skills. Algorithms are inhuman forces that bring social, political, material and cultural formations into being, generating and extinguishing possibilities. Their inhumanism transmutes ideas of the human and demands new (post)humanisms.

This issue of Transformations presents contributions that address the inhuman algorithm.

The rapid development of predictive technologies from simple pre-emptive text to voice-activated virtual assistants raises questions about how we engage with bodies of knowledge mediated by algorithms. Predictive technologies with increasingly adaptive algorithms supported by machine learning, have the capacity to learn alongside us, gleaning information to better understand behavioural patterns and predict human action and intention. These technologies are often promoted in terms of how they assist human users and are evaluated in terms of their speed and relevance. This valorisation of speed is underpinned by an algorithmic means-end logic that is not subject to the durational constraints of human perception and attention. Indeed, the inhuman time of an algorithm has to be adjusted to fit the lived time of human thought and action. Drawing on the work of Henri Bergson and Bernard Stiegler among others, this paper argues the quest for speed in the development of search technologies constructs a future in which time is reduced to discrete possibilities and disregards the lived delay immanent to human thought.

Keywords
Pre-emption, Internet Search, voice assistants, predictive technologies, time, thought, cognition

The use of machine learning and algorithmic processes to screen for mental illnesses, and to propose potential treatments, has caused some to be concerned about the possibility of increasingly impersonal and invasive forms of psychiatric surveillance. For some, the rise of big data and algorithmic psychiatry presents the possibility of a future where the mentally ill are increasingly dominated by machines. Concerns about the use of algorithms in the diagnosis of mental illnesses, and the devising of treatments, perhaps overlooks the extent to which an algorithmic revolution has been facilitated by existing human-centred psychological therapies. Through the dominance of cognitive-behavioural therapy (CBT), commonly experienced mental illnesses such as anxiety or depression have come to be understood as the result of faulty processes of recursive thinking. The model of consciousness that underpins cognitive-behavioural approaches, suggests that to be human is to already engage in forms of cognition that are open to algorithmic manipulation, insofar as to think is to produce recursive rules of self-conduct. Accordingly, this paper seeks to articulate the inhuman model of thought that is assumed by CBT, and to consider how it has opened the space for an algorithmic revolution in mental health.

Keywords
CBT, Algorithmic Thinking, Biopolitics, Therapeutic Culture, Mental Illness

In this paper I suggest the idea of algorithmic superstructuring as a way to explore aesthetic regimes of algorithmic governance, drawing on work of Jacques Rancière, Luciana Parisi and Wendy Chun. Algorithmic superstructuring presents as pervasive expansion of algorithmic processing and logic, installed under the techno-capitalist drive for quantifying, consolidating and regulating human experience. Algorithmic superstructuring is built into networks of distribution and circulation of affect and flourishes in the cognitive frameworks of interfaces and protocols. Building on previous curatorial work and drawing on media art practices, this paper aims to investigate how inhumanity of algorithmic modes and models of reasoning is reflected in the distribution of the sensible, and how the aesthetic regimes of algorithmic governance could be articulated.

Keywords
algorithmic superstructuring, media art, algorithmic governance, distribution of the sensible

Algorithmic modes of thought have long and problematic histories of collusion in processes of governmentality, dating at least back to the Atlantic slave trade and including the othering of neurodiverse, black and indigenous, and queer cultures. But beyond their instrumentation within systems of power, this paper proposes that at the foundation level of algorithmic design there are a series of assumptions about what constitutes legitimate thought processes. These assumptions are based on neurotypical modes of thought and often ignore the possibilities of more neurodiverse thinking, which is regularly devalued in our society. This naturalised “whiteness” that lies at the centre of and colonises algorithmic programming needs to be interrogated and rethought, it is argued, in order to break the relationships between algorithms and oppressive power systems. Drawing on fugitive and devalued modes of thought such as queer kinship and failure, black sociality and the incomputability at the heart of the mathematical concept of Omega, the article speculates on the conception of a minor algorithmic value or “life” closer to that of an emergent collective and ecological consciousness than that of the dominant individualised and fixed model that is valued within contemporary capitalism.

Keywords
Neurodiversity, Algorithms, Queer Theory, Sociality, Governmentality, Self- Organisation, Brian Massumi, Jack Halberstam, Fred Moten, Omega, Luciana Parisi, Black Studies