Page 2 – Foreword – Gina Neff, Minderoo Centre for
Technology and Democracy
Page 3 - Take Back the Sense – Anna Berti
Suman, Tillburg University
Page 4 - Blue Solutions to Greening the Internet – Anne Pasek, Trent University and Hunter Vaughan, University of
Colorado
Page 5 - A Responsibility Void: - Federica Lucivero, University of Oxford
Page 6 - Big Tech Goes
Green(washing) – Joana Varon, Founder Directress, Coding Rights Fellow on Human Rights and
Technologies, Harvard Kennedy School and Camila Nobrega,
Founder, Beyond the Green PhD Candidate, Free University of Berlin Fellow,
The New Fellowship
Page 7 - Driving Headless into Autonomy – Helen Stamp,
University of Western Australia, Julia Powles, University of Western Australia
and Audrey Guinchard, University of Essex
Page 8 - Carceral Systems Will Not Save Us from the Climate
Crisis – Mallika Balakrishnan, University of Cambridge and Julia Slupska, University of Oxford
Page 9 - Beyond Big Tech’s Epistemic Imperialism – Jessica
de Jesus de Pinho Pinhal, Technische Universität Berlin
Page 10 - How Can We Democratize Decision-making
on Data Centre Construction? – Julia Rone, Minderoo
Centre for Technology and Democracy, University of Cambridge
Page 11 - Digital Participatory Technologies and the
Environmental Impact of ‘the Extractive Gaze’ - Dr Brenda McNally, School
of Architecture, Planning and Environmental Policy, University College Dublin –
UCD
Page 12 - Entanglement -
Fiona McDermott, Trinity College Dublin
Page 13 - AI is perpetuating
climate injustice – Dr Theodora Dryer, AI Now, New York University
Page 14 - Commoning: An alternative governance paradigm for our
digital futures – Deborah Thomas, Fellow, Young Leaders in Tech Policy,
University of Chicago & Foundation for Ecological Security
Please cite this
report as:
Minderoo Centre for Technology & Democracy. The Cost of Convenience.
September 2021.
We reached out
across several academic fields to challenge an international group of experts
with
the task of writing a provocation on the impact that
digital technology has on the environment, pushing otherwise careful academics
to enter into speculation, critique, exchange, and
dialogue in order to bring transparency to this
opaque issue.
We quickly
discovered a need for bridging across several distinct and siloed conversations
and across the problems of media infrastructures,
the tech sector’s environmental impact and the potential harms of our digital
information environment our natural environments. We found a need to scope new
research on tech’s environmental impact and a
need for collaborative and international efforts to identify areas that require immediate action and research. What you are
reading is the first step in this process. Like the ‘zines’ of the Punk Rock era that inspired our report’s style, we
also see our work as a provocation for you to join us in continuing.
These provocations are a fitting
start for us at the Minderoo Centre for Technology and Democracy,
because we see our main goal as supporting the democratic values in
everyday digital technologies. As a centre, we challenge assumptions about power in two senses. The first, as we and
our international collaborators do
here, is to think about the material consequences of our digital lives - how
electricity and resources and infrastructures
of digital technologies are implicated in the choices that we make as every
day.
Four key common
themes emerged across the provocations that we present here. The first set of
provocations concern the extraction of data and resources needed to power
technology-saturated lives. The second cover the costs incurred without a
‘right to repair’ - the planned obsolescence of the digital age
that encourages people to consume without regard to costs over the lifecycle of
digital products. Third is the challenge posed by greenwashing, including the
prevalent narratives about digital technologies as weightless, clean and green and the companies that produce them as
blameless in the environmental crisis. Finally, these provocations question the
digital exhaust and the resources and infrastructures needed to maintain it.
What links these
four themes and the provocations that follow is that they all suggest how
citizens and communities can question the values and choices of technology
companies through a lens of power. That brings us
to the second sense of power that we at the Minderoo Centre for Technology and
Democracy are committed to: restoring the balance of social and political power
to citizens in light
of the increased dominance of large companies making choices about data-driven
infrastructures
in our everyday lives.
We convened the
Cost of Convenience Workshop in June 2021 when much of the world was still
grappling with COVID-19 restrictions. What
emerged was a model for collaborative thinking and the first steps in new connections to help us reclaim that collective voice,
articulating what counts over what is convenient, what is valuable over what those in power value.
We hope that we can
inspire you, too, to begin new ways of challenging the cost of convenience with
us.
Gina Neff, Minderoo
Centre for Technology and Democracy
Citizen sensing,
environmental monitoring from the grassroots with the use of our own senses or
sensors, can be regarded as a practice that aims at appropriating technology
for the sake of health and environmental protection.
However, when it
comes to using sensors to monitor the environment, it is a contradiction to use
tools produced by the tech industry which is one of the main causes of those
adverse environmental effects that people fight.
Civic actors try
to resist to such mainstream technologies by building their own frugal
equipment to track environmental degradation otherwise.
I will discuss
how the auto-production of technologies is a needed counterforce, although
still not enough.
I will defend
that we need to ‘take back the sense’ and focus on using our bare senses (e.g. sight, smell…) to detect environmental contamination as
only this way we will truly reconnect with nature,
without
contributing to further waste production.
Let’s talk about network infrastructure...
Many of the provocations in this zine
are concerned about the environmental impacts of the Internet. But the
different parts of the Internet are not equal in impact and benefit.
Our research group focuses on the
subsea cables that carry over 95% of our global commerce and communications.
They’re an essential, but small, part of the ICT sector. Their carbon footprint
is also really marginal, so much so that they’re often omitted
from attempts to quantify the climate impacts of
digital networks, or seen as something
of a rounding error.
As such, scholars and policy makers
who work on the problem of the tech sector and climate change generally ignore subsea
cables. Why look for solutions where there doesn’t seem to be a problem?
Because sometimes problems - and solutions - exist
deep below the surface.
Our provocation is to change this, asking how taking
cables into account might help us imagine different and greener forms of
network connection.
A shift to moving more
data over the subsea, rather than duplicating and storing it in an extensive
system of terrestrial data centres and content delivery nodes, would be a win
for the climate. What’s more, it might also be a win for the communities in
which cable landing stations are built: small island nations and coastal
municipalities facing uncertain energy futures and increasing rates of sea
level rise. We are currently investigating how these global infrastructures
could be leveraged to provide local benefits to public health and economic
development for historically marginalized populations by shifting to 100%
renewable power. Engagement at this level of the Internet could also provide
avenues for lessening the global digital divide.
In short, to focus on the hidden
parts of our global networks is to focus on local geographies, community-based
approaches, and alternative networks.
We find much hope and many
opportunities at this scale.
Despite the
promises of enabling sustainable development goals, the infrastructures that
make digital services, data production and processing and training of AI models
possible have a relevant environmental impact throughout their lifecycle. This
is due to CO2 emissions, deployment of natural resources for hardware
manufacturing and emissions of toxic substances during the disposal of
electronic devices.
Who should be
accountable for this digital pollution?
Shall I watch
less streamed videos? Delete my photos and send less emails? Yes, perhaps, but
responsibility cannot be placed uniquely on individual citizens as they are
only one element of a larger ecosystem that should enable sustainable
behaviours.
The larger
ecosystem, however, does not seem to offer reliable accountability instruments.
Currently, there are very few regulatory
tools, as policy makers seem to hold to the hope that
digital technologies will evolve to the point of reducing their own
environmental footprint. Action is left in the hands of the private sector, where companies agree on codes of
conduct and self-regulatory tools that are not binding, very diverse,
and not adhering to common standards.
Some stakeholders
in the field don’t find this fluidity helpful as the lack of guidance implies
that some industries (for example the ones developing infrastructures) need to
do all the work while others (the ones developing digital services) are not
considered as liable.
How to fill this
responsibility void? My contention is that to fill the void we need to start by
rethinking the concept of responsibility that we use in this context.
Responsibilities
for technology-related impacts are traditionally understood in terms of
the liability model, where one person can
legally be made to pay for the adverse effects of their actions on
others. However, the climate change discussion on responsibility often moves away from this individual-focused and
retrospective model of responsibility to stress the need to adopt a collective
approach that takes into account the distributed
dimension of environmental actions and allows a prospective look at its
implications.
My question is, can we reconcile these framings in thinking about the sustainability of digital
technologies?
Over the last
years, following the buzz around the Green Economy, most big tech companies
have made a series of commitments to reduce carbon emissions, all widely
disseminated in marketing campaigns.
In parallel, they
have also positioned themselves as the suppliers of almost magic “techno solutions”
to solve climate change. Once they were gonna save
democracies... now they will save the Earth?
Our take is that
Big Tech are going green (washing), profiting from a perfect match of Green
Economy and Technosolutionist narratives. Therefore we bring feminist lenses to unveil the connection
between
both discourses, and their role in recent theoretical frameworks.
Dangerous
discourses that erases damages.
For instance,
Alphabet have suppliers extracting minerals from mines devastating the Amazon forest and bordering indigenous land. What other
information do we have about social-environmental
conflicts caused by big tech businesses?
From
mineral to data extractivism, what power relations
are being imposed in particular territories? Which
inequalities are being reinforced? How can we build a decolonial and non-extractivist digital society?
The tech and
automotive industry pitch that autonomous vehicles deliver a safer, more
efficient future than current cars is so pervasive
that it has become an unquestioned touchstone for policymakers. But what is the
true environmental impact of the intensive computation and data processing
required to develop and use autonomous vehicles?
When fleets of
autonomous vehicle are trialled over countless miles to earn our trust,
what is the scale of the computational demands and ongoing strains on digital
infrastructure? How does this compare with a human driver’s emissions? And, most importantly, how is this impact weighed against
the ever-alluring goals of increased vehicle safety and driver convenience,
promoted by tech firms and endorsed politically?
Let’s do a quick
addition on just some of the hidden environmental costs of autonomous
convenience:
·
Energy and infrastructure needed to sustain network connectivity, transmit
data, and render maps and offer computer vision across vast geographies. One of
the most interesting and neglected dimensions of this is the subsea
telecommunications cable network (see work by Anne Pasek
(Trent University) and Hunter Vaughan (University of Colorado));
·
Energy needed to run data centres (see work by Mél
Hogan (University of Calgary) and Julia Rone
(University of Cambridge));
·
Water-intensive needs in order to build the
various computer chips in both vehicles and data centres (see work by Janna Z
Huang (University of California, Berkeley))
·
Metal extraction costs of the elements needed to build electronic
circuits, leading to destruction of tropical forests (see work by Joana Varon
(Coding Rights) and Camila Nobrega (Free University of
Berlin));
·
Carbon emission costs to get components from where they are extracted to
where they will be used. Similarly,
the absence of rights of repair drives waste and further extraction (see work
by Steve Jackson (Cornell University), Maio Lu (The Chinese University of Hong
Kong), and Jack Qui and Ziyi Wang (National
University of Singapore)).
Now complicate
the picture further. What if our democratic requirements for autonomous
vehicles don’t reduce these demands on the Environment, but instead compound
them?
What if it turns
out that data retention is necessary in order to
investigate and determine issues of fault and liability when autonomous
vehicles inevitably crash, and the uncertainty around applicable rules
incentives a culture of ‘collecting everything’?
Some
technologies, most particularly autonomous weapon systems, provoke a sharp
normative question: should we build them? A quick addition on the environmental
consequences of autonomous vehicles suggests that if we want better futures, we
should at least ask the same question. Should we?
Environmental activists disrupt the convenience that
powers climate change. Whose convenience has been a source of tension in
the environmental movement: the past years have seen a shift from disrupting
individual consumers (e.g. by blocking roads to promote
a message to drivers) to corporate and state decision-makers. In 2019,
thousands of Silicon Valley workers went on strike in protest of their employers’ climate impact,
directly targeting tech companies (1).
Silicon Valley positions itself as having the tools
to save the planet, but its business model of extraction and accumulation are
part of the problem. Further, its technologies facilitate the surveillance of environmental
activists, including through contracts with law enforcement. Carceral systems
will not produce the tools to save us from climate crisis.
Digital platforms prompt activists to share data,
offering a variety of affordances that help activists mobilise and connect with
others. In doing so, these platforms also make activists more legible to law
enforcement. In India, Google, Facebook, and Zoom seemed to cooperate with the government’s arrest of climate activist
Disha Ravi for sedition (2). In the US, Facebook turned over detailed records on the indigenous-led protests
against the Dakota Access and Keystone XL pipelines (including messages to and
from the group’s page and a list of everyone “invited” to the protest event)
(3).
Surveillance of protest movements both during and in
between protests takes up a much larger share of police time and resources
than is commonly
understood (4). In the UK, undercover police
officers posed as members of environmental justice groups and initiated
romantic relationships with women, in one case even fathering a child while assuming a false
identity (5). UK police also placed Extinction Rebellion (XR) on a list of
extremist groups that should be reported to Prevent. Even so, groups like XR
which are predominantly white and middle class, and have a history of
cooperating with police, are likely to be treated much better than people of
colour doing this work. Surveillance tech and environmental destruction target
racialised communities hardest.
We must investigate the cost of convenience in
terms of convenient platforms, convenient contracts between tech and the state,
and convenient ideas about clean, green tech that produce billions for a few while entrapping the rest of us in racial surveillance
capitalism.
White-washed climate activism and green-washed tech
obscure the need for systems change towards sustainability, as opposed to quick
fixes. Rather than treating tech companies as saviors,
we should hold them accountable for extending carceral systems that harm our
best shot at environmental justice. We look to movement-led structural change
that tackles environmental destruction and its links to broader systems of
oppression-- that means no surveillance tech in our climate justice.
1)
Matsakis, Louise. 2019. “Thousands of Tech Workers Join
Global Climate Change Strike.” Wired.
2)
Klein, Naomi. 2021. “India Targets Climate Activists With The Help of Big Tech.” The Intercept.
3)
Davis-Cohen, Simon. 2018. “The Justice
Department Helped a County Prosecutor Target the Facebook Records of
Anti-Pipeline Activists” The
Intercept.
4)
Gillham, P.F. 2011. “Securitizing America: Strategic Incapacitation and
the Policing of Protest Since the 11 September
2001 Terrorist Attacks.” Sociology Compass, 5: 636- 652 -
https://onlinelibrary.wiley.com/doi/10.1111/j.1751-9020.2011.00394.x
5)
Lewis, Paul and Rob Evans.
2020. “Secrets and lies:
untangling the UK 'spy cops' scandal”.
The Guardian
The realm of ubiquitous
computing has a singularity; it spreads everywhere but is nowhere to be seen.
Hooked on our
screens and the digital worlds they display, we
overlook the devices themselves.
We choose to
ignore the resources, both human (data labellers, Uber Eats drivers) and
material (rare metals, silicon), necessary to the magic at play.
The
digitalisation of the product and the fragmented globalisation of its
production lead to the uncontrollable growth of Big Tech. I argue that the only
sustainable and desirable futures must draw on theories of degrowth.
But they are
often limited to an economic, ecological, or post-colonial framework.
I propose
epistemology to frame the dichotomy between growth and degrowth within the
well-known cultural, epistemological, and theological dialectic of
universalism/particularism, objectivism/subjectivism, and
transcendence/immanence.
Big Tech’s epistemic imperialism is
nothing else than a modern and technological continuation of the perilous
project of Man: the domination of both Nature and the Other.
As more and more
forms of communication become mediated by digital tech, the need for storing
and processing data dramatically increases. Thus, it is no wonder that in the
last few years we have seen a boom in the construction of data centres whose
high electricity and water demand is a stark reminder of the “costs of
convenience”.
Drawing on my
work on local resistance to data centre construction in the Netherlands, I
argue that decision-making about data centre construction has such important
environmental, social and economic impacts that it
cannot be left to big private companies and government executives alone (at any
level of government) but should be much more democratic. The question I want to
pose is then: How can we democratize decision-making on data centre
construction?
This is not an
abstract, theoretical question. Exactly the opposite: it was triggered by my
analysis of the discourse and grievances of citizens of the Dutch province of
North Holland who were angry not only about the environmental consequences of
data centres but also because no one had even bothered to ask them for their
opinion. Rather than being against data centres per se, citizens of North
Holland argued that they wanted more information, more ways to participate and
to discuss whether/how data centres could work for the benefit of their
communities.
To be sure, we
still need to figure out whether the demands of building a global
communications infrastructure by global corporations could take
into account factors such as community well-being. Maybe yes. Or maybe
this is a lost cause and we should rather think of
ways to scale-down and build a communications infrastructure that is more
decentralized, interoperable
and context-bound.
In any case,
finding how to balance our communication needs with not destroying the
environment and local livelihoods requires a decision-making process that
includes more voices, more involvement of national parliaments but also of
municipal councillors and “ordinary” citizens. The current practice of
behind-the-curtains negotiations between big private corporations and
executives (be they at the national or local level)
is increasingly contested.
Ultimately,
democratic participation is both normatively and practically desirable. If we
want to think seriously about the environmental impacts of tech and how to
mitigate them, we need to involve the very local communities that experience
first-hand the impacts of data centres. Informing citizens and their
representatives about digital infrastructure and guaranteeing ways for them to
participate is not utopian. It is exactly how democracy should function.
Our daily lives are dominated by images. Today, we live
in, and through, digital media which are made consumer-friendly by prioritising
the visual content. Within this context, questions about what is made visible
and how seeing, knowing and power are related, are essential to understanding
the impacts of digital media on contemporary society. However, these questions
also apply to the use ofdigital participatory
technologies which are fundamentally changing our way of seeing and
encountering nature.
Citizen engagement with the climate and biodiversity
crisis is considered essential to enabling the radical social transformations
required to address these challenges in a democratic way. These engagement
processes are increasingly mediated, drawing on a range of digital
participatory technologies, such as drone technology, AR/VR, geolocation
platforms and visual apps to emotionally and experientially
engage citizens through geometric abstraction of nature, place and
landscape.
While government and local authority actors understandably
focus on the efficiency and effectiveness of these tools compared to
traditional engagement initiatives, we must also ask how digital tools are reshaping socio-ecological relations. For example, drone
technology is increasingly deployed to engage citizens via film, documentary
and advertising with nature and the environment.
However, these dramatic images also represent nature as
geometric abstractions, thereby visually separating people from nature, and
normalising existing unsustainable nature/society hierarchies.
Similarly, critics of participatory mapping technologies
argue that these new ways of seeing radically change our experience of
geographic knowledge from the God’s-eye view of the map to the embedded
subjectivity of GPS. In doing so, these visually spectacular renderings of
nature subordinate the land to human purposes, perpetuating and justifying
existing power relations. This raises the question, what are the environmental
impacts of this ‘extractive gaze’ and cartographication?
As digital participatory technologies increasingly dominate citizen engagement
initiatives with environmental challenges, we need to ask: what is being
prioritised and marginalised in these new nature/society encounters and with
what future environmental impacts?
Most
significantly, how do we balance the need for increased, democratised
participation, with the associated costs of data intensification and the
normalisation of unsustainable socio-ecological relations?
This provocation to consider the cost
of convenience of data technologies, comes in
the form of a physical exhibition entitled Entanglement, as curated by the
interdisciplinary design, art and research group, ANNEX.
The exhibition
sets out to dispel the myth of the internet as an ethereal or
abstract entity and instead provokes visceral understandings of
how data production and consumption territorialise
the physical environment. The exhibition argues that the digital
is material - the cloud is a complex, spatio-temporal reality that has distinct material and
environmental consequences.
In particular, Entanglement examines the position of Ireland in the
evolution of global data infrastructures, both historically and into the present day. Today, the materiality of
the data economy reveals itself locally across the Irish landscapes as a vast
network of data centres, energy grids and subsea cable landing sites. in 2019,
Dublin overtook London as the data centre capital of Europe. By the year 2027,
data centres are forecast to consume thirty-one percent of Ireland’s total
electricity demand.
The exhibition
uses the prism of heat, to foreground the thermodynamic processes necessary for
data production, storage and distribution. Foremostly,
the pavilion asserts that from the burning of campfires to the management of
waste heat generated by contemporary data processes, the production and
dissemination of information is intrinsically connected to the production and
dissemination of heat.
Entanglement questions if any single
discipline can understand and meaningfully
respond to the environmental challenges brought about by the exponential growth of data technologies.
The complexities of these new
territories require interdisciplinary connections
between the humanities and engineering, as well as the bringing together of
those from industry, activism and academia.
Artificial
intelligence developed in the name of benefiting the environment is not the
same thing as establishing environmentally and socially conscious AI systems.
AI is a powerful
terminology advanced by technological developers
that can refer to a broad suite of technologies, data collection and storage
infrastructures, and automated decision systems. It often
describes predictive statistics-based digital systems such as machine learning.
Ascertaining whether or not this variegated typology of technological
systems can be used to address the climate crisis foremost requires
identification of locally situated technologies and data, as well
as the environments they extract data from and transform through subsequent AI-backed
decision-making procedures that determine environmental policy and resource access.
This is
especially true in the domain of water rights and water access. Digital
automated decision systems are currently in prolific use in water allocation,
distribution, and diversion policies. In these contexts, the water data fed into these systems streams from a longer history of information extraction and control that is inseparable from the natural resource policies and legal
apparatus that it functions within. Extant conditions of inequity and
coloniality at work in these contexts are often hidden by the technical
function and opacity of the digital systems.
It is therefore
imperative to center justice and sovereignty rights,
rather than economic growth agendas, in assessments of AI and environmental
policy. Furthermore, by clarifying the terms - of what exactly, is meant by
‘artificial intelligence’ - it is evident that promises made that AI will
‘solve’ the climate crisis are in direct opposition to how AI functions in
perpetuating climate injustice.
While
commoning involves collective action by communities
and people groups; the institutions, governance frameworks, and rules required
for such action are based on systems of knowledge. It is these systems of
knowledge that shape environmental discourses and resultant
action at nested levels of governance, both of the community
as well as the state.
I posit that commoning
of technology (data, information and digital systems)
is essential to building technologies and
systems of knowledge that are environmentally and socially just.
The idea of commoning, as an alternative socio-economic and governance
paradigm, provides a framework for actors at
nested levels of governance to self-govern and determine the future of tech.
It refers to the
participatory governance mechanisms that we should build around data,
information/knowledge and digital systems. Since each
of these constituents of digital tech have their own logics and political
economies, it is necessary to delineate what commoning
means for each.
Environmentally
just data involves countering the extractive logic that separates data from
provenance (of people and lands) and reinforces power hierarchies.
Power is embedded in each stage of the data lifecycle, and hence environmental
data justice would mean that questions of access, production, infrastructure,
governance are to be asked at each of
these stages.
A Commons framework
which is premised on ideas of self-governance and collective action, mean that
communities determine what data is digitized, ensure that data stories
(provenance and metadata) are ground-truthed, and can
mobilize data to challenge environment policy that is premised on extractive
logics. It can also lead to data infrastructure and production methods that are
participatory, equitable, and transparent.
Going by Fritz Machlup, information, and subsequently knowledge are built over the foundation laid by data (or, raw facts).
While it is necessary that this information is not digitally barricaded, the
social systems that assimilate this information into knowledge and enable its
use in decision making, are equally crucial. The evolution of these social
systems, in ways that embed considerations of local livelihoods and ecology,
can be fostered by the participatory processes that commoning
entails.
Finally, digital
systems themselves, both hardware and software, must also be ‘commoned’. Both design and technology management practices
are central to ensuring that tech does not cause unintended harm
; and that the knowledge systems intermediated by tech reflects
pluralism.
Common forums
between technologists, their end users, and other stakeholders, leading to
collective action that debates visions of social good, prevents centralization
and shifts power, could provide avenues where the true costs of convenience are
judged.