[Air-L] Reminder: last chance: Please join us for the AoIR 2020 Pre-conference: Algorithms and the future of agency

[Jakob Linaa]

Dear all

So far, a whopping 165 participants have signed up for our
AoIR pre-conference: "Algorithms and the future of agency" . There are
still some vacant online seats before we reach our limit. I will close for
sign ups tomorrow.

The event will take place online Monday October 26th at 17.00 UTC.


Sign up here: https://forms.gle/sS64W8MXX8ej3AsQ8

Zoomlink to be provided to participants the day before

Descriptions and program below.

Organizers are: Jakob Linaa Jensen, Stefania Milan, Mark Andrejevic, Dylan
Wittkover, Andrew Iliadis, Emiliano Treré, Aline Shakti Franzke.




Idea:
As more and more daily and social life is mediated through platforms,
algorithms play an increasing role determining our information and capacity
for action. For instance, the increasing use of automated decision-making,
by insurance companies, hospitals, police and government in general raises
important ethical and philosophical questions on the future of agency.
First, in algorithmic procedures and decision-making, it is difficult to
ascribe agency and thereby responsibility. Second, if our independent
agency, also known as the free will, is severely limited by algorithmic
decision-making and algorithmically defined horizons of information (for
instance on Facebook and Google), it bears important and far-reaching
consequences for liberal democracy as we know it. Responsibility,
accountability and agency are fundamental prerequisites of democracy. If it
is impossible to identify power and keep certain actors accountable,
democracy and individual rights are at stake.

In this pre-conference, we invite participants to explore and discuss such
issues. To initiate and facilitate the discussions, a range of scholars
with different background and interests, but all interested and engaged in
the topic, will provide igniting talks discussing aspects of algorithmic
agency. The first four talks sketch the problems from political,
democratic, semantic and ethical perspectives. The last three talks sketch
possible solutions, both conceptual and practical. We hope by the talk to
inspire discussions based on specific experiences with algorithmic agency
and to provoke suggestions for possible solutions in order to save agency
and, ultimately, life and democracy as we know it in an increasingly
algorithmic world.


Program

In the first half we will discuss the problems of algorithmic agendy and
the first four organizers will give igniting talks (app. 10 minutes each
followed by questions and comments from the audience). In the second half,
we will discuss possible solutions to the problems, based on the igniting
talks by the last three organisers. At the end of the session we will try
to draw a map of these issues and questions that might form a backbone of a
coming joint publication for interested participants.

PART I – The problems of algorithmic agency

Political agency in the age of AI: The case of facial recognition tech
Stefania Milan

How does artificial intelligence, and facial recognition technology in
particular, alter the exercise of political agency? Taking the perspective
of ordinary citizens, the contribution will touch upon two key aspects of
democratic life, namely political participation and the relation between
the state and its citizens. Briefly teasing out the possibilities and the
perils for active citizenship entrenched in the expansion of AI into
society, this contribution will raise questions on the future of democracy
in relation to a technology which will have unparalleled consequences on
human life.

Semantic Agency: Graphing Knowledge on Web Infrastructures
Andrew Iliadis

Agency is tied to the power to name, classify, categorize, and organize. In
today’s web infrastructures, knowledge about entities and relationships
that exist in the world is codified into semantic web code/syntax that
undergird countless information products, from Google knowledge panels to
the Wikidata that feeds into virtual assistants. Such graphing of knowledge
impacts areas as diverse as news consumption, fact checking, advertising,
and logistics. This contribution draws on recent theoretical work on the
philosophy and politics of web semantics (Monea 2016; Hui 2016; Thomas
2018) while also drawing on empirical research and case studies that focus
on the work of platform companies and organizations who build and maintain
semantic infrastructures on the internet.

Algorithmic agency and the future of democracy
Jakob Linaa Jensen

Agency is closely related to the power we willingly or unwillingly impose
towards ourselves in the platform economy, for instance when we sacrifice
privacy for convenience. But such power is blurred and difficult to
identify. It is a fundamental democratic principle that actors can be held
responsible for their actions. If this is not possible, it has far-reaching
consequences for liberal democracy as we know it.

Ethical challenges of algorithmic agency
Aline Shakti Franzke

Agency has been a key term of ethics since Aristotle. In order to hold
something accountable classical definitions have ever since focused on
intentionality and mental representation. These human centered definitions,
however, are stressed when applied to algorithms. Therefore, new ethical
challenges occur along three lines a) meta-ethical concerns b) the agency
of algorithms themselves and c) the implications of algorithms for human
agency. The talk will map existing ethical debates along these axes.

PART II – Possible solutions

>From autonomy to respectful interdependence
Dylan Wittkower

Postphenomenological analysis is less frequently used by Internet
researchers, and offers distinctive contributions to an understanding of
the future of agency. Postphenomenology avoids the reductive extremes of
technological neutrality or technological determinism by offering
descriptions of the lifeworld as constituted by hybridized human-technology
agents (in embodiment technics) that interact with hybridized
technology-world objects (in hermeneutic technics). I argue that this
postphenomenological description of technological mediation encourages us
to seek the future of agency not in autonomy or control, but in in
respectful interdependence as articulated by feminist ethics of care.

How “smart” technologies bypass subjective agency
Mark Andrejevic

Recent Foucault-inflected work on "environmentality" explores the ways in
which environmental level monitoring and modulation (as envisioned in some
versions of the "smart" city, for example) can operate as a strategy for
control that bypasses subjective agency. The so-called libertarian
paternalism of "nudge" economics, for example, might be described as a
popularized form of environmentality that manipulates behavior below the
threshold of reflexive consciousness. Such forms of manipulation come to
rely on increasingly comprehensive forms of data capture designed to
generate correlational predictions about future behavior under modulated
circumstances. These developments emerge alongside ongoing attempts to
bypass mediation by tapping straight into the human brain (as in the case,
for example, of Elon Musk's Neuralink). I'll talk about possible
alternatives and challenges to such developments.

Resisting algorithms
Emiliano Treré & Tiziano Bonini

In this talk, we propose a conceptual framework to map the practices that
citizens perform to shape, interfere and resist algorithms. Mobilizing the
concept of the moral economy, we first situate the various forms of
algorithmic agency along a continuum shaped by the two competing moral
economies of the users and the platforms. Secondly, we bring into the
equation the type of power held by those who enact specific forms of
algorithmic resistance, adding the strategic vs. tactical poles. Finally,
we discuss our framework drawing on examples including data obfuscation,
streaming frauds, audience boosting, rating improvement, profile
optimization, visibility enhancing, engagement groups, etc.