More selected projects

Canvassing the API

Essenzia: Intricately Personal Modern Art - proven by data

"Who are you?
You’ve always wanted to see. We can find the essence of you to be displayed as your own unique and special work of art.
We use the power of data to craft an artwork which is truly meaningful to you, aesthetically pleasing especially to you - and stylishly modern.
Simply answer our short set of questions, expertly devised to provide an accurate insight into who you really are."

Essenzia is a speculative company demonstrating worst practice in the zone of intersection between capitalism, data analysis and art.

Website at (use password "essenzia")

produced by: Hazel Ryan


This project attempts to trace and problematise what can be termed the fetishization of data 
science in the arts, specifically more or less abstracted forms of data visualisation within
computationally-produced or situated contemporary western visual arts - although this is far
from the only location of such fetishization. Indeed I touch on tracing this issue back to some of
the routes of development leading to the contemporary western art market. The term
‘fetishization’ is strong, but it evokes the ‘data is sexy’ view propounded from the 2010s
onwards (Davenport and Patil 2012), and deliberately situates the reverence of data in the arts
as potentially harmful.

I first examine a conceptual framework through which I attempt to theoretically situate data-led
artworks, and by which it might be possible to describe examples of harmful or of progressive
practice. I am careful to avoid a full taxonomy of ‘good’ and ‘bad’ practice, instead pointing
out traits of artworks which might be regarded critically.

My practice-led piece builds on the critical work of artists such as Morehshin Allahyari & Daniel
Rourke to speculate on an extreme example of unregulated corporate data practice whose
alternative truth is lent credibility by the wrapper of art. I freely make a critical value judgement
against the embrace of data fetishization by the arts from a postcolonial and .

I argue that through unchecked celebration of data science techniques, art does the work of
capital by lending aesthetic credibility to often self-referential pseudoscience, and as a
consequence, colonial and surveillance-focused business and political interests. Art is treated
as ‘bare art’ (Sholette 2018) in this context, hoped to be imbued with an immunity to political
criticism, and I explore this phenomenon further.

I state strongly that I am not referring to all use of data in art - when data is pure medium, with
no agenda of persuasion behind it, the results can be striking. Art can also perform data
advocacy through actively seeking to highlight data injustice (of collection, processing or
application), and I explore forms of resistance in this light towards the end of the report.

Conceptual Framework

If we are to investigate the effects of working with data in art, then it is first necessary to look at
some of the theory which might lend a critical eye to this. Data visualisation can present
numerical knowledge in ways which aid understanding and ease of engagement - Tufte. More
widely, data-led statistical practice is a rich field useful to many social applications,
internally-consistent and scientifically rigorous. However privileging this form of knowledge as
superior to others would be a kind of platonic realism or positivism.

Bruno Latour’s actor-network theory alludes to data as thing-in-the-world, implying there is a
materiality and method to data visualisation and it can’t be interpreted as just abstraction or
neutral quantification or measurement - for example the visualisation itself of financial market 
price data, with color and chart, along with the data content, influences further actions of
traders (Nafus 2016, 385-386).

At the same time, data visualisation is a topological method in artistic practice, allowing artists
to access processes and ideas that wouldn’t otherwise be possible - but also leaving the
practitioner and the viewer open to the problems of twisting, stretching and questions of
witnessing that distance brings (Awan 2016).

Following a discussion on research and creative practice, including gathering data through
surveys and focus groups, Linda Candy suggests [Interacting 49] that “instead of simply
appropriating research methods from other disciplines and contexts, an alternative approach is
to find out what best matches artistic aims and knowledge and the creative process itself”.
Here she is referring to research-led creative practice whereby potential audiences are involved
in the creative process, but I see this as easily extendable to the wider data gathering process
for data-led artworks. If art is to be interpretable as a form of knowledge different to data
science practices, its content cannot be predicated on the methods of data science.
○ We cannot see the world in a new light through data practice because our
knowledge is what has defined the discipline. e.g. “Semantics derived
automatically from language corpora contain human-like biases”
%3A0%3A0%3AkWcyn86OnJQ2%2BVYKnm7lBg%3D%3D) (Kiritchenko and
Mohammad 2018, 44).

● Does the art world want to borrow from data science when there exists, for instance, a
gender data gap in every examined discipline, whereby women are relegated to
navigating a world built around male data? (Criado-Perez 2019, XI-XII).
● Queering data practice - how do we ensure that art does not simply reproduce the
exclusion of the experience of marginalised groups that happens, for instance, in big
data practice in the business world? And how do we resist complicity in the use of
personal data to further stigmatise/pathologise marginalised people such as LGBTQ+
people? ( Gieseking 2018, 8).
● Six Provocations for Big Data - “Claims to Objectivity
and Accuracy are Misleading” (4-5).

Describing Data-Led Art Practices

I discuss here some examples of data being fetishized in art through visualisation - a common
theme is mystification of the collection or analysis process and of the translation from data to
form, borrowing from tools developed by and for business, or in direct collaboration with these
creators of business tools; and then the final work being presented as objectively illustrating
some measurable ideal, while being too abstract in form to actually convey this type of
knowledge. That is, where the work is lent credibility through its aesthetic appeal, and then
becomes describable in any way which suits the purpose of the creator or owner, following a
pseudo-positivism but exempting itself from the rigours of positivist instruments of proof by
lurking under the banner of art.

At its root, this tendency might be seen as an example of the human inclination to shore up
existing interests or viewpoints by using affirmatory science (O’Neil 2016, 48). But a rush to
embrace new forms of quantification is at the root of many poor data practices in business. Big
data algorithms and social sentiment analysis engines alike encode human prejudice,
misunderstanding and bias into a black box “beyond dispute or appeal” (O’Neil 2016, 3).

Data visualisation can be a rich aesthetic practice (see, for instance, Giorgia Lupi and Stefanie
Posavec’s Dear Data project (Lupi and Posavec 2015)), but I see it important to be wary and
critical of referring to it as an art in its own right, as to do so removes it from a critical lens and
risks placing the practice in a category where it might be exempt from the rigour of scientific
practice by placing the emphasis foremost on its aesthetic appeal, or affording it the gaslit
reverence reserved for non-platonic knowledge by western establishments - meanwhile being
used as justification for business or political decisions in the way that scientific study might be.

First I looked at Golan Levin’s The Dumpster from 2006, an example of pure visualisation
situated firmly in the art world (Whitney 2006). This had the fantastic premise of plainly
presenting the romantic lives of American teenagers. However on examination, the data is
heavily selective, based on the shadowed wisdom of a corporate third party, and mysteriously,
extensively processed. The project attempts to distill rich human experiences into quantifiable
numeric facts, using proxies, or stand-in metrics, to measure what it is really interested in. This
process is described by Cathy O’Neil (O’Neil 2016, 17-18, 145-6) as inevitably slipping away
from the question of actual interest, and this is a fundamental problem for many attempts to
quantify experience. In this case the question of interest is “How similar were these breakups?”
but the proxies used are instead the emotional ‘feel’ of a post, the inferred age of the writer,
and so on.

○ Worrying sentences: “Using custom language-analysis software, the text of each
post was computationally evaluated in order to determine many different
characteristics of the breakup and the just-ended relationship.” This could mean
almost anything. Language-analysis software has severe limitations and certainly
did in 2005 [SOURCE]. Intelliseek was a “marketing intelligence firm that
provides technology-driven solutions to measure and analyze online
word-of-mouth behavior for marketing and advertising professionals”
med-500-List-Fastest-Growing-Private-Companies] [footnote for next bit?] What
is it about Intelliseek which appealed particularly to the artist in processing the
most intimate information of privacy-unsavvy teenagers? No statement is given
on how this sensitive, sometimes identifying information was processed or
stored by the company, or which third parties might have had access to the
extrapolations made from the blog metadata.
○ Kamal Nigam has expertise in data mining and machine learning with an
emphasis on analyzing text and internet data. Formerly Director of Applied
Research at Intelliseek, he has just started a position at the new Google
engineering office in Pittsburgh [he now runs this office].
○ Andrea Boykowycz developed an efficient database tool for post hoc data
cleaning. Many thanks to her, and to Jessica Greenfield, Joel Kraut, and Giana
Gambino for their assistance in pruning the dataset. - what does this mean!?

These artworks in themselves are not necessarily enacting significant cultural or social harm, but
they contribute to the culture of fetishization as ‘microaggressions’ of legitimisation which
ultimately pave the way for an acceptance of corporately-developed data visualisation methods
as true art practice and extrapolating the value of art-knowledge to apply to these tools. When
this becomes of critical importance is when these methods in art practice explicitly intersect
with capitalism, and I will describe this effect below.

Towards the end of the report, I will also contrast these examples with examples of resistance
through data-led art, whereby artists are using carefully chosen datasets, not over-extrapolating
or confounding their significance, and perhaps borrowing tools from business but directing the
gaze of these tools back on the practices of those holding structural power in data.

Contextual Review

Artists who are directly addressing overuse, misuse or fetishization of data in the arts include
Morehshin Allahyari & Daniel Rourke, who created a compendium of 3D printing activist
interventions, and Mimi Onuoha, who “seeks to explore the ways in which people are
abstracted, represented, and classified” through data.

As direct context, it is also productive to consider some of the most egregious examples of
data fetishization occurring in the periphery of the art world. This region is what I have focused
on in creating an artefact. I’m particularly fascinated by seeing that the offers of data-driven art
studios, advertising houses and analytics startups are becoming more aligned; harder to tell 
one from the other as they both use a reverence of data as a pivotal tool in the fight for
credibility through aesthetics.,, (note the
parallel uses of the .ai and .io domains),,,,,, (^1)

Machine learning/AI are often seen as ‘magical’ rather than coming from ‘dry’ statistical
methods, which plays into this approach - sometimes to the point where straight commentary
is hard to distinguish from satire, such as this article entitled “An AI god will emerge by 2042
and write its own bible. Will you worship it?” (Brandon 2017). In fact, machine learning
algorithms were not only recently invented and big datasets have been analysed by
statisticians for decades: “many of the methods and techniques we’re using [...] are part of the
evolution of everything that’s come before” (O’Neil and Schutt 2013, 2).

(^1) TEKJA stands out in another way among these examples, profiting both from capitalist interests and
their resistances, and leveraging their work as art to enable this profit from both cause and solution and
perhaps distance themselves from responsibility over their choice of client base through an implicit
appeal to Sholette’s ‘bare art’. Their London Data Streams work, as displayed in the Big Bang Data
exhibition at Somerset House in 2016, was positioned firmly under data visualisation-as-art (BBC
specifically describes Big Bang Data, and especially this piece, as art Surely this caught the eye AON Insurance, who
commissioned TEKJA in 2017 to create an ‘interactive microsite’ for The AON Global Risk Management
Survey’ to pinpoint the greatest risks in the global industries they specialise in
( While AON provide risk-management and ‘cost
reduction opportunities’ for the defence, energy and mining sectors
s-Industry.pdf), TEKJA also helped their clients Forensic Architecture and Amnesty International to
expose the details of Israel’s 2014 Gaza offensive
(, critically
dependent on the global defence sector which its other client AON specialises in supporting [too
conspiratorial?]. Their client Novartis ( - selected clients) just gained US FDA approval
for the most expensive patented drug in the world
( while the viz company also
claims the NHS as a client
(, increasingly
unstabilised by privatising interests, and directly battling Novartis’s interests in some cases


Analysis of my own project

I have touched on the works of theorists and practitioners above who address the phenomenon
of data fetishization in more or less circumspect ways. Through the creation of my artefact and
the speculative intervention around this, I hope to extend this work, while drawing critique to
what I see as poorly justified data practices such as those in the works outlined above. Prior
entwinement of ‘art’ and data has paved the way for probing the boundaries of what might be
passed off as legitimate art practice, in a way which in fact purely serves capital by charging the
user for the collection, storage, processing and transfer of their data, its subsequent
obfuscation and rendering into an artwork. The speculative corporate interest I have created,
Essenzia, is designed to be particularly egregious through its demand for more identifying data
than might be considered necessary; its worrying privacy policy, as well as its allusion to
scientific process with no actual respect of research standards.

The misleading, almost meaningless methodologies presented by Essenzia
( - password “essenzia”) are partially
drawn from real examples of jargon and unverifiable methods presented in data practices. All
the survey questions are hyper-stylised examples of the problematic proxy metrics used in data
collection which skirt around the measures of actual interest where these are not directly
observable (O’Neil 2016, 17-18, 145-6). Question 4, “You see a person eating an apple with a
knife and fork. How do you respond?” is indeed a real question found on an online quiz to
determine political ideology, at Sentiment analysis, as loosely performed
through Question 5, “Choose the sentence that appeals most to you.”, is coldly applied to
blocks of text on a word-by-word basis with biases between speakers from certain
demographics (Kiritchenko and Mohammad 2018, 44). Requesting residential property type, as
in Question 6, parodies the postcode metrics and focus on housing estates/projects used to
criminalise minority communities (O’Neil 2016, 86-88).
Form gives legitimacy to data - the mapping process to create the artefact is nonsensical,
statistically unjustifiable - if there’s any objectivity to the end result, it is only that it says nothing
about that which it claims to describe. Yet just through being told that the object means
something, it begins to take on those qualities. I know that the data mapping process was not
chosen in any structurally meaningful way, at least under the terms which it presents itself, yet
by spending time with the object it has begun to reflect back something about myself, through
aesthetic appeal and a sort of bond formed through sensing it. This has given rise to an
unexpected postphenomenological reading of the artefact - the object has become a
“thing-for-me”, subject and object inseparable through experience (Rosenberger and Verbeek
2015, 11). This ties in with Latour from before. It is interesting to contrast these flavours of
postphenomenology with Kirby’s Quantum Anthropology, in which she accepts that data
visualisation may actually illustrate something objective, despite twisting and changing

Data-led art under capitalism

If they are not succeeding in providing a clear elucidation of a phenomenon through scientific
rigour, but they have written themselves off, through this pretence at objectivity, as existing
solely as art objects, then what purpose are works such as those examples above serving? With
an eye to my own practice as described above as a particularly ugly distillation, I argue that
they are serving capital rather than serving social benefit or aesthetic ends.

The tools behind such works are developed in corporate settings, and are often full of biases
and under-examined in their usefulness or ethics (Berman, Pekelis and Van den Bulte 2018).
Co-opting these for art is taking for granted that these tools can be used to paint a useful
picture of the world; as evidence of the universal world - this is the case, for instance, in Levin’s
The Dumpster, uncritically using the tools of Intelliseek, whose methodologies were unclear.

Borrowing these tools concurrently places art practice into the toolbox of corporate interests to
gloss over their practice. Especially when displayed in environments lending credence to
corporations, “a work of art also suggests a cultural authority, a form of dignity, even of
wisdom, which is superior to any vulgar material interest” (Berger 2008, 135). An egregiously
transparent example of this motive comes in the statement by Robert Kingsley of Exxon, as
quoted by Hans Haacke in his 1975 work On Social Grease (Haacke 1982): “Exxon’s support of
the arts serves as a social lubricant. And if business is to continue in big cities, it needs a more
lubricated environment.” Business has turned its attention in the present-day to data-led art,
which not only provides this lubricant through aesthetic validation, but cleverly legitimises the
tools of business analytics as (hidden) subject.

An example of art lubricating the work of capital via data visualisation is the “ERROR - The Art
of Imperfection” exhibition at the Volkswagen Group Forum in 2018-2019 (Ars Electronica
2019). Data visualisations were described as art, and presented in a gallery-style layout in a 
multi-purpose space that also served as a luxury car showroom (^2).The whole field of curation 
tells us that the situation of presentation of art matters immensely. An artwork is fundamentally
tethered to an institution which provides a platform for it. The idea that art might exist in
isolation rather than being altered or even created by the exhibition space leads to Gregory
Sholette’s concept of “bare art” (Sholette 2017, 66).

(^2) The Ars Electronica website gushes, “In the Berlin DRIVE. Volkswagen Group Forum it’s all
about deviations from the norm.” (Ars Electronica 2019). Do these deviations include emissions
deviations as exposed in the 2015 scandal? (Reuters 2017)


While he focuses on the art market itself - profits made from the 1% of art sales which are
ultra-high-value - and authoritarian regimes’ aggressive accumulation of cultural capital, such as
the UAE Culture Summit, I wish to apply this term more widely to apply to the view that art is
art regardless of its situ; art can be valued in its own right regardless of the circumstances
leading to its creation; that must be employed in order to value works of pseudo-pedagogical
data visualisation as works of art, or particularly as works of art which lend credibility to their
context. This is a wilful ignoring of the developmental history of art under capitalism and/or to
serve capital. The context might lend credibility to the art; where the opposite is achieved a
critical investigation might be performed of who or which institution stands to benefit from the
display or sale of the artwork. If, as Berger asserted, much art in history has been based on the
desire to take possession of the objects they depict (Berger 2008, 83-4), then who is set to
acquire the business tools implicitly depicted by data-led artworks?

Maybe this has been an inevitable development in fields describable as computational art
practice, as digital practices, when reduced to their components and processes, are all
sculpted out of data in some sense, and so computationally-produced data artworks are in
some sense self-referential by nature. As early as Robert Rauschenberg’s 1966 live data piece
Open Score (Medienkunstnetz n.d.), as soon as art incorporated data visualisation it was taken
by the spectacle of data itself. The 1968 Cybernetic Serendipity exhibition at the ICA was
beholden to the process of computing itself as implicitly interesting to visualise (Monoskop
2018), and perhaps a joy in new tools themselves is a right that artists can claim as technology

However this focus on the spectacle of data in art becomes what can be termed a colonisation
of art by data at the point at which an artwork engaging with data is attempting to provide
some objective truth on the world, or persuade to some point of view beneficial or important
to the creator or exhibitor. Nick Couldry and Ulises Mejias describe a “data colonialism” in big
data practice in particular, in which “social life all over the globe becomes an ‘open’ resource
for extraction that is somehow ‘just there’ for capital.” (Couldry and Majias 2018, 1-2). I see this
as extending to any collection and processing of the data of those holding less power by those
with more power, as performed by data analysis companies whose tools are used by artists,
such as Intelliseek as above. As artist Geraldine Juárez states as part of her work “PROSPEKT”,
drawing parallels between modern corporate data capture and seventeenth-century colonial
bio-prospecting, “existence is affirmed through perpetual capture: nature into culture into
data. All into capital. Organising information is never innocent.” (Juárez, 2018).

Art which glorifies borrowed quantitative content serves to further deprivilege naive art,
outsider art, and any art lying outside a formal institutionally-recognised set of training, by
platforming art predicated on some form of academic or corporate practice. Much as data
gathering and representation have been democratised, these methods still stem from
mathematics departments and business labs. Are we seeking Mahmoud Sabri’s Quantum
Realism, describing “the application of the scientific method in the field of art” (Sabri 1975, 77)
to transfer “the artistic function from the level of appearance to the level of essence”?


The artists I mentioned in my contextual review, whose work I have been directly inspired by,
are in some sense all performing acts of resistance against uncritical data narratives in art and
beyond. Beyond this specific focus, other acts of resistance in the arts serve to disrupt the
complacency of data fetishization.

Looking beyond direct visualisation, the We Need Us project interests itself in crowdsourced
data classification, dealing in metadata instead of data itself (We Need Us 2018). This still
results in highly visually ambiguous results, but there’s a decentering of the ‘power of data’
narrative, looking at the “specific qualities of open data and metadata once any ‘useful’
information that can be analysed and put to conclusive use has been discounted”. They also
take the important step of conceding “we don't know exactly what we will find out about data
through the work”, an exploration rather than a colonisation of meaning.

On a more explicitly political level, the artist Hasan Elahi, having been the subject of FBI
tracking, chooses to directly resist surveillance through creating more tracking data about
himself to ‘flood the market’ and devalue that which the government can collect (ACM
SIGGRAPH n.d.). The data offered directly by Elahi includes the high-volume mundane 
snapshots of the type routinely collected by social media companies and offered up for analysis
and visualisation.

I have so far tacitly supposed that an answer to the problem of art serving capital through data
is by less widespread adoption in the arts of tools of data capture and processing. But perhaps
a weakening of the view of legitimacy of these tools can also come through a less co-operative
borrowing - through flooding the market with junk data, deliberately misusing the tools (as I
touch on in my practice) or through turning the lens of the tools on the practices of their
creators. Perhaps a direction for further resistance could be a modern answer to Hans Haacke’s
piece “Shapolsky et al Manhattan Real Estate Holdings, A Real Time Social System as of May 1,
1971”. Haacke gathered data on real estate developments and transactions over two decades,
public records critical to their business dealings, and directly exposed fraudulent activity
(Whitney n.d). Could we use Google’s data analytics tools to shine a new light on the
company’s data-colonial practices?


My argument can be summarised as a critique of art under capitalism as performing the work of
wealth accumulation, not principally through the sale of works, but through the public
legitimisation of the corporate process. This extends to grants and funding, ostensible support
provided by the private sector which of course shapes the direction taken by artists’ work. This
is nothing new, but by focusing in particular on data-visualisation or analysis-led pieces
presented as art, I am illustrating a second layer to this parasitic relationship, in which not only
does the artwork give credence to capital interest, but also the data tools themselves used to
create the artwork, often the same tools as used by business, are legitimised and promoted to
embed this relationship further than traditionally.

To attempt to guide the theology of statistics to bear fruit in the medium of art is to do a grand
disservice to the tactile, emotional (more theoretical terms!!) knowledge provided by art. It
seeks to impose a realist idea, of an ontologically independent truth ready to be revealed by
quantitative methods, onto a discipline which has thrived precisely on its separate, parallel
tradition of knowledge which values subjectivity and impression as valid and illuminating. I will
venture to say that art does not need statistics. Art needs data as pure tactile phenomenon,
procurer of interesting patterns, of pixels on screen or marks on canvas. Art does not need data
as its authority on the world. It does not need to buy into the same trappings of windowed
objectivity, the same conflicts of origin and interest. If it ventures out from the same starting
point as does science and science’s (peripheral, xenomorphic) offshoots (quantitative politics,
data science, marketing) then it becomes one of these subfields, taking part in the same
overconfident post-rationalisation fueling all the world’s transient vessels of capital collection
[quote NNT on epistemological arrogance?], and thus unable to say anything about them due
to being entrenched by the same forces.

Problems to address
● Why am I holding art to ethical standards?
● There is a continuum between misleading data practice in art and good data practice

Further questions and developments
● How might art and scientific practice work together outside capitalist interest? Beyond
the scope of the resistance I’ve mentioned.

Website at (use password "essenzia")


ACM SIGGRAPH. (n.d.). Tracking Transience. [online] Available at: [Accessed 10 Aug. 2019].

Ars Electronica. (2019). ERROR - The Art of Imperfection. [online] Available at: [Accessed 10 Aug. 2019].

Awan, N. (2016). Digital Narratives and Witnessing: The Ethics of Engaging with Places at a Distance. GeoHumanities, 2(2), pp.311-330.

Berger, J. (2008). Ways of seeing. London: British Broadcasting Corporation.

Berman, R., Pekelis, L., Scott, A. and Van den Bulte, C. (2018). p-Hacking and False Discovery in A/B Testing. SSRN Electronic Journal.

Brandon, J. (2017). An AI god will emerge by 2042 and write its own bible. Will you worship it?. [online] VentureBeat. Available at: [Accessed 12 Aug. 2019].

Couldry, N. and Mejias, U. (2018). Data colonialism: rethinking big data’s relation to the contemporary subject. Television and New Media, vol. 20, 4: pp. 336-349.

Davenport, T. and Patil, D. (2012). Data Scientist: The Sexiest Job of the 21st Century. [online] Harvard Business Review. Available at: [Accessed 10 Aug. 2019].

Haacke, H. (1982). On Social Grease. Art Journal, vol. 42, no. 2, pp. 137–143.

Juárez, G. (2018). PROSPEKT - Geraldine Juárez. [online] Available at: [Accessed 10 Aug. 2019].

Kirby, V. (2011). Quantum Anthropologies. Durham and London: Duke University Press.

Kiritchenko, S. and Mohammad, S. M. (2018). Examining Gender and Race Bias in Two Hundred Sentiment Analysis Systems. Proceedings of the 7th Joint Conference on  Lexical and Computational Semantics, pp. 43-53.

Lupi, G. and Posavec, S. (2015). THE PROJECT — Dear Data. [online] Dear Data. Available at: [Accessed 10 Aug. 2019].

Medienkunstnetz (n.d.). Media Art Net | Rauschenberg, Robert: Open Score. [online] Available at: [Accessed 10 Aug. 2019]. (2018). Cybernetic Serendipity - Monoskop. [online] Available at: [Accessed 10 Aug. 2019].

NAFUS, D. (2016). The Domestication of Data: Why Embracing Digital Data Means Embracing Bigger Questions. Ethnographic Praxis in Industry Conference Proceedings, 2016(1), pp.384-399.

O'Neil, C. and Schutt, R. (2013). Doing data science. Sebastopol: O'Reilly Media.

O'Neil, C. (2016). Weapons of Math Destruction. London: Penguin Books.

Reuters. (2017). TIMELINE - Volkswagen's long road to a U.S. Dieselgate settlement. [online] Available at: [Accessed 10 Aug. 2019].

Rosenberger, Robert and Verbeek, Peter-Paul (2015). Postphenomenological Investigations: Essays on Human-Technology Relations. Lanham, MD: Lexington Books.

Sabri, M. (1975). Quantum realism. Baghdad: Alvataniah.

Sholette, G. (2017). Delirium and Resistance. London: Pluto Press. (2018). We Need Us. [online] Available at: [Accessed 10 Aug. 2019]. (2006). The Dumpster. [online] Available at: [Accessed 10 Aug. 2019]. (n.d.). Shapolsky et al. Manhattan Real Estate Holdings, a Real-Time Social System, as of May 1, 1971. [online] Available at: [Accessed 10 Aug. 2019].