On Pandemonium: a new work by Ami Clarke

Do Androids Dream of Pandemonium, available as an 'online dashboard' here, was commissioned by Radar for Risk Related and developed further during Clarke's residency at ZKM | Center for Art and Media Karlsruhe in the summer of 2021. It forms part of Clarke's ongoing body of work exploring probability and risk within surveillance/disaster capitalism.

Written in February 2021, the below essay by Clarke serves as an introduction to some of the issues explored by the work. 

Risk in times of crisis: hybrid strategies.

“[T]he coronavirus is us: we live in an interconnected world, where borders are porous, more like living membranes than physical walls” – Michael Marder

The events of the past year have a persistently quoted condition of being ‘unprecedented,’ but it’s clear that pandemics are anything but un-anticipated, as made clear by the government's test drill ‘Exercise Cygnus’, which showed the woeful extent to which Britain was underprepared for a pandemic.

The pre-pandemic moment crystallized around wealth derived from extractive relations that inscribe ongoing inequalities, as the ‘magic sauce’ of memetic media met the ‘secret sauce’ of right-wing billionaires in underwriting political campaigns to facilitate a wholesale move to the hard right in the face of the upcoming environmental challenges ahead. Epidemiologists now advise government economic strategies, as existing inequalities come to the fore fast, facilitated by an unprecedented inter-connectivity that reaches across a neoliberal globalised workplace, the effects of which remain inflected by specific geographies and their socio-economic materialities.  A state of contingency—a chaos of churning markets and mediasphere—provides maximum opportunity “to exact ‘emergency measures’ with minimum scrutiny for oversight - in the pandemic version of ‘disaster capitalism’”.

In previous works I focussed on how the past is dragged into the future via the contractual conditions of both the derivative, and insurance, with the persistent problem of anticipating the unforeseeable, within the realm of probable outcomes. The events of the last year will have unfortunately forced many to become considerably more familiar with the shortcomings of insurance policies meant to cover the disastrous outcomes of ‘unprecedented’ events, but which have tended, thus far, not to cover the effects of ‘novel’ viruses such as SARS-CoV-2[1] This has left many unable to claim or facing steeply increased costs (existing policies for care homes, at one point over the past year, were claimed to have increased by as much as 800%).

Yet, as noted above, pandemics are anything but ‘unprecedented’. Angela Mitropoulos’ work shows that contractual conditions, such as insurance, and other forms of financialisation that come of the neoliberal paradigm, date a relationship with contagion considerably pre-Covid19. 

The history of insurance, with its associated risk management and probability modelling, shows the increased impact of ‘unprecedented’ events: from insurers refusing cover to those in flood or fire zones, to the new financial instruments built to accommodate more frequent events in the form of parametric insurance, primarily in response to environmentally related catastrophes. Typically, models reliant on probability depend on past data such as HURDAT datasets, drawn from historical hurricane analysis in order to predict future events. But limits to modelling catastrophe come to a head in the uncomfortable fit between stochastic models used in catastrophe modelling and the increasingly dynamic modelling of climate science. Now, ‘let it rain’ insurance policies go beyond prediction tethered to past events to deploy real time sensors which provide immediate payouts, once a specific hazard parameter is met.[2]

Meanwhile, virus hunters working with re-insurance firms draw attention to the increasing dynamic in animal-to-human viral transmissions, brought about in part through the increased capacity for animal human entanglements resulting from deforestation, for example, with the additional capacity to spread across the world, exponentially, due to the increased connectivities brought about by a globalised workplace; an unprecedented interconnectivity that foregoes boundaries, with unforeseen consequences, as the failure of nation states to safeguard their own populations against austerity leads to a reinforcement of national identity and a weaponization of xenophobic tendencies in a return to populist politics. Attempts to ‘manage’ the flow of infection via the track and trace apps further blur boundaries, revealing complex interdependencies between the app’s unprecedented surveillance—designed to manage risk and modulate the flow of bodies within a viral environment—the analysis of which informs states and other players, as well as the markets.  

My previous work The Underlying focussed on how interdependencies found across social media, insurance and the market impact upon the carbon cycle at the heart of the climate crisis, as the reputation economy of social media proves to be a driving force in the markets. As public opinion turns, so too does the insurance industry's willingness to underwrite the future of the fossil fuel industry - rendering oil, gas and coal as toxic assets that subsequently need to stay in the ground; a particularly stark example of how probabilism operates in measures of ‘degrees of belief’, often understood in terms of what one is prepared to lay wager on, as a measure of conviction itself.

The pandemic also reveals many interdependencies, of course, but following the thread of probabilistic means of measuring conviction, with a slightly different onus.

Sentiment Analysis

The epidemiologist and risk modeler Nita Madhav was at a family wedding on December 31, 2019. As recounted by Evan Ratliff in Wired, she received texts from Ben Oppenheim, her colleague at Metabiota, a company which specialises in advising governments on epidemics, informing her of a cluster of unusual pneumonia-like infections in Wuhan. They had been flagged by the firm’s early detection system, which included an algorithm for parsing and highlighting news stories about outbreaks in similar ways to sentiment analysis, measuring differing degrees of sentiment to discern vital clues to what was happening on the ground. The team typically examined hundreds of media reports a week and approached new ones cautiously. At the reception, Madhav messaged with Oppenheim and wondered: if it was respiratory, could the source be more like H7N9, the avian flu? A coronavirus like SARS-CoV? 

Within weeks, Madhav was living inside Metabiota’s own projections. Specializing in contagious disease monitoring by utilising data from a network of research stations to conduct disease surveillance for clients (which had previously contributed to the USAID project ‘Predict’: ‘a database cataloging viruses in their animal reservoirs in order to forecast which might jump to humans’, which had been closed by the Trump administration in September 2019[3]), she and her team studied historical data on thousands of major disease outbreaks since the 1918 flu, to create what they called an ‘Epidemic Preparedness Index’, an assessment of 188 countries' capacity to respond to outbreaks. As Ratliff recounts, the group also began “tracking and studying how media coverage evolved around different types of outbreaks. Scarier diseases tended to generate more news stories … In 2015 the Zika virus outbreak arrived and crystallized the reality that fear was a critical variable in understanding the economics of outbreaks.”  The “economic consequences of a scourge, the historical data showed, were as much a result of society's response as they were to the virus itself.” Ratliff goes on to describe how the group start building what became known as the Sentiment Index, informed by University of Oregon Psychology Professor Paul Slovics research, on how human beings perceive and respond to risk. They gathered their own information from around the world on how much various symptoms frightened people. Fear, understandably, had as much to do with the economic fallout of a pandemic as governmental policies.

Munich Re and Metabiota’s Sentiment Index and Epidemic Preparedness Index considers a country’s or company’s resilience to a pandemic, or other “natural” disaster, in order to produce a ‘Financial Vaccine’.  They include the health capacities, but also “non-health system factors” such as “financing, institutional capacity and infrastructure”- drawing together surveillance, immunization, physical infrastructures such as roads and internet, as well as political stability and levels of corruption, health spending as well as public health communication capacities - in 188 countries. Ksenia Chmutina and Jason von Meding’s work points towards how language related to disaster reporting impacts upon what is understood about how ‘disasters’ become ‘disastrous’, by ignoring these contributing factors of a human sort.  Warning against this, they quote the United Nations Strategy for International Disaster Reduction, which defines a disaster as “a serious disruption of the functioning of a community or a society at any scale due to hazardous events interacting with conditions of exposure, vulnerability and capacity, leading to one or more of the following: human, material, economic and environmental losses and impacts”. Despite this, they note that “many scientific disciplines refer to disasters as ‘natural,’ with only the social science disciplines’ research epistemology, more aligned with a social construction lens able to reflect “how this expression sits uncomfortably at best, particularly given the contemporary understanding of the role of vulnerability in driving disaster impacts on society.” The importance of the “viral chatter” identified by Metabiota and others, in places where animals and human boundaries blur and entangle, was key to a further understanding of how the serious implications of environmental damage and pollution, were driving the likelihood of further pandemics, that could far surpass the current one. In The Underlying I was keen to point to how socio-economic factors mean that pollution is often underpinned by economic inequalities and poverty, with tax evasion an important and ongoing financial extraction and oppression, often in places most needing of important infrastructural development.  As Chmutina and von Meding note, “hazards can turn into a disaster because of human acts of omission and commission rather than an act of nature, and [that] disasters are caused more by socioeconomic than natural factors.” 

Surveillance / disaster capitalism.

A state of contingency that provides maximum opportunity “to exact ‘emergency measures’ with minimum scrutiny for oversight - in the pandemic version of ‘disaster capitalism’”.

The track and trace apps quite naturally exceed all previous anxieties regarding surveillance, reaching across biomedical practice to behavioural analysis, and, at the same time, hold the promise of a pragmatic approach, which, seemingly, when used in combination with measures such as mask wearing, social distancing protocols and an effective vaccination programme, could mean the possibility of a return to being able to co-mingle, once again – as can be seen in several countries success in controlling the pandemic, the world over.

But the history of surveillance runs deep, emerging in tandem with the capabilities of the various technologies of the time, to measure, define and scrutinise bodies, in several distinct but overlapping histories. Science in the service of colonialism and capitalism, from the medical-industrial complex to the pharmaco-pornographic[4] era, has long been put to work inscribing racial difference, gender as a reductive binary relation, sexual heteronormativity and the overwhelming privileging of able bodies, with platform/surveillance capitalism pre-empted in Foucault’s conception of the panopticon as internalised self-surveillance.  Numerous recent advances in technology, point to the legacy of these histories, drawing out like a poultice existing biases and discriminations, precisely because of a self-serving and erroneous notion of tech ‘neutrality’ – a tabula rasa of tech, if you like – which is as much a fantasy as the notion of ‘a level playing field’.  For example, the US' National Institute of Standards and Technology took until 2019 to find that there was a one hundred times more likely chance of people of colour being misidentified through facial recognition technology than white people, with Sen. Ron Wyden adding that the findings showed how “algorithms often carry all the biases and failures of human employees, but with even less judgment.”[5]  Humans (increasingly in combination with machine learning amid an array of cognitive and non-cognitive devices) design, interact and engage with software, through very specific dynamics, that include the highly exploitative practices of capitalism, often predicated upon extractive relations that can be traced to colonial times.  Legacies of this sort also continue in the disproportionately higher death toll of people of colour from Covid19, where the economic realities of discrimination, link employment opportunities with the option to even work from home at all.  It should come as no surprise that anxieties relating to any or all of these histories might accompany a new app intended to moderate behaviour in the seemingly innocent pursuit of managing the rate of infection in a pandemic.

Concerns that accompanied the roll out of the first apps test period on the Isle of Wight tended towards whether Bluetooth signals would work sufficiently well to achieve accurate enough results or were robust enough to prevent data seepage, to considerations focussed on the anonymity of data held in either centralised, or locally distributed databases, with varying means by which to certify that the data was only used for the purposes it was intended for.[6]  The strategies regarding the build of the app, though, bely the larger and far more insidious aspect of mission creep, not only for the purpose of purloining the new oil of data, but through the project’s financial underpinning, with strategies not designed to help, but intended to undermine the public health service that it is seemingly in support of. Oscar Williams writes that the news of Palantir’s involvement in developing the Covid-19 Datastore (alongside Faculty, Amazon, Google and Microsoft) attracted criticism from privacy advocates when leaked in January 2021, with speculation about the company’s expectations of securing further work with the health service at the end of the project, and the lucrative revenue stream it could generate for the company. Notably, Palantir, known best for its work with the US government in powering ICE immigration raids, the defense sector, and police surveillance, is part of a group of companies that include Serco, Deloitte, Sitel, G4S and others, tasked to manage the test, track and trace programme, some of whom have entirely dubious past records for this kind of work.  Labour MP Diane Abbott (Radio 4 news bulletin Feb 2021) points pertinently to how the notorious ‘hostile environment’ in part facilitated by companies such as GS4, significantly contributes to a deserved lack of trust in government, that now extends to GP’s, leading to serious problems of uptake in both the vaccination programme and of course anything to do with ‘track and trace’.  Richard Brooks writes that “the same firm now entrusted with tracing Covid contacts, Serco, has in the past few years admitted fraud and false accounting in contracts for offender-tagging. In a leaked email, Serco’s chief executive said he wanted the scheme to “cement the position of the private sector” in the NHS supply chain.[7]  Its auditor and now partner on test and trace, Deloitte, was also fined £4.2m for signing off Serco’s fiddled numbers. Meanwhile, the effect of outsourcing to private companies can be best seen in the probation services where the damage was so great that the job had to be brought back ‘in-house” last year, the chief probation inspector judging it ‘irredeemably flawed’.”[8] 

The ideological combines with so-called pragmatic concerns in the concentrated onslaught of austerity cuts and outsourcing, which has meant the NHS has simply not received adequate funding for well over a decade and more.  And this continues during the pandemic, with openDemocracy reporting that Medacs Healthcare plc - a subsidiary of Impellam Group, a FTSE-listed firm whose largest shareholder is Lord Michael Ashcroft, the Belize-based Conservative peer who has donated millions to the Conservative party (one of the ten largest donors), being awarded a £350million contract to recruit staff for the vaccination programme, when the NHS could dearly do with that kind of support for its already functioning recruitment programme — with numerous other instances of this sort.[9]  Peter Geoghegan and Russell Scott, continue, referring to a highly critical National Audit Office report that found that “companies with political links were directed to a “VIP lane” for government contracts where bids were ten times more likely to be successful.  Shadow cabinet office minister Rachel Reeves told openDemocracy: "People are understandably furious seeing businesses owned and run by the friends and donors of the Tory Party being awarded huge multi-million-pound public contracts throughout this pandemic. “Cronyism, incompetence and waste have been everyday features of this government's approach to outsourcing and ministers show little willingness to learn lessons from the National Audit Office investigations."” Rather than ‘levelling up’ the instinct has been to ‘shore up the private sector’ in a pandemic version of “disaster capitalism” writes Rachel Shabi in 'The Pro-Privatization Shock Therapy of the UK’s Covid Response going on to describe how “Britain’s public health sector, a cash-strapped, eroded, but functioning network comprising the National Health Service (NHS), general-practice clinics, and local authority health officials, has been repeatedly sidelined in favour of outsourced alternatives. As Allyson Pollock, a professor of public health at Newcastle University, put it to The New York Times recently: “They’re basically trying to build a centralized, parallel, privatized system.”

Shabi spoke to Tan Dhesi, a Labour member who, back then, reminded Parliament in mid-June, that the government had suspended normal procurement rules and a competitive bidding process during the pandemic and went on to describe a slew of private-sector pandemic contracts as purely “ideological.” In this case, it’s not even “privatization through the back door,” he said—the pillaging is taking place in full view.  The casual disregard of due process in procurement has resulted in criticism from global commentators, as well as members of their own party, who note a catalogue of reckless misspending leading to missed opportunities in managing the pandemic, that plagues the current government. Out of exasperation at the still not functioning track and trace system, several councils set up their own systems, utilising local knowledge, that have thus far proven considerably more effective. [10]

When Shabi wrote her article in July 2020, the death toll from the pandemic in the UK had passed what seemed a frightening number of 44,000. Today the figure is unforgivably over 100,000. You could understand how, perhaps, one of the strongest metrics in the Epidemic Preparedness Index: that of ‘trust’ in governments, could be seen as left wanting. 

Since writing, the government has doubled down on its ‘private takeover’ under the slogan: “taking back control” of our NHS.  A proposal, hurriedly revealed after a leaked report from a thinktank critical of the NHS Covid response was shown to have made donations to the health secretary Matt Hancock.[11] The government quickly announced plans to reinvigorate the NHS - and solidify its ambitions in allowing private companies like Deloitte to sit on boards that make decisions about how to spend NHS money. 

Pillaging in full sight, as boundaries blur, seeping, across enmeshed ecologies from animal to human to mink to further mutation. 

To be continued.


[1] The coronavirus pandemic has led to widespread disruption and business closures resulting in substantial financial loss. Many customers have made claims for these losses under their BI insurance policies. There has been widespread concern about the lack of clarity and certainty for some customers making these claims, and the basis on which some firms are making decisions in relation to claims.

The issues surrounding BI policies are complex and it was recognised that they had the potential to create ongoing uncertainty for both customers and firms. The FCA accordingly sought clarification from the High Court as part of a test case, aimed at resolving the contractual uncertainty around the validity of many BI claims. The FCA advanced arguments on behalf of policyholders in the public interest. The test case was based on a representative sample of policy wordings. The High Court’s decision on the test case was subject to a leapfrog appeal to the Supreme Court.

The Supreme Court handed down its judgment on 15 January 2021 substantially allowing the FCA’s appeals and dismissing the insurers’ appeals. This means that many thousands of policyholders who have cover should now have their claims for coronavirus-related business interruption losses paid.

The judgment from the Supreme Court and High Court provide policyholders and insurers with clarity about whether customers have cover and can make a valid claim and the amount due to policyholders. See here for further information. 

[2] Until recently, individual consumers didn’t have access to parametric insurance in the US. That’s changing: in October 2018, a company called Jumpstart started offering earthquake coverage to Californians. The trigger is a quake that reaches 30 centimeters per second of peak ground velocity, a measure the US Geological Survey uses to create “shake maps” of intensity. This article in Wired details this development. More recently, parametric insurance utilising data from sensing technologies, “such as the seismometer network in California or the hurricane-hardened WeatherFlow anemometer stations on the East Coast” has opened up possibilities for new ways of thinking about insurance: “…you can get that data and very quickly work out whether someone should be getting paid,” says Samuel Jay Gibson, of the Capital and Resilience Solutions Group at the catastrophe risk modeling firm RMS. “This allows post-event, initial injections of cash for immediate disaster recovery.”

[3] This included work in collaboration with the Wuhan Institute of Virology.

[4] Paul B Preciado’s terminology. See Testo Junkie and other writing.

[5] The National Institute of Standards and Technology, the federal laboratory known as NIST that develops standards for new technology in the US, found in 2019 what it calls “empirical evidence”, that most facial-recognition algorithms exhibit “demographic differentials” that can worsen their accuracy based on a person’s age, gender or race[vii] with the likelihood of people of colour being misidentified up to 100 times more likely than white people. https://www.washingtonpost.com/technology/2019/12/19/federal-study-confirms-racial-bias-many-facial-recognition-systems-casts-doubt-their-expanding-use/

[6] Steve Clark of IT consultants Clark Associates wrote a well-circulated article, which noted that as the trial started there was a lot of debate on social media “variously combining the view from the Information Commissioner’s Office, the view from the Parliamentary Joint Committee on Human Rights and the open letter signed by 170+ academics many with privacy and information security within their purview.” The Joint Committee also addressed human rights, noting that “the lockdown measures are a significant interference with the right to family and private life” before noting that digital contact tracing would to ease the lockdown, but that it in “itself will interfere with the right to private life.”

[7] See here

[8] A study of outsourcing last year by the Institute for Government found that, while contracting more routine services such as catering and waste collection from private companies appears effective, “government should be cautious about extending outsourcing of frontline services”.” Brooks goes on to explain, “the arithmetic of outsourcing explains much of the problem. To make it worthwhile, public bodies typically look for savings of around 20 to 30% of existing costs. The private company also needs its own double-digit profit margin, so the cost at which it provides the service needs to be far lower than it was in the public sector. Lower wages and skills and thus poorer service follow.” And not just in the short term, Brooks writes, “such narrow calculations in any case miss the longer-term costs foisted on the taxpayer: dealing with the increased reoffending from failed probation and prison rehabilitation, for example; or picking up the pieces from what MPs called the “fiasco” of outsourced disability benefit assessments or an “abysmal” deal for Capita to run military recruitment.” 

[9] “A healthcare company ultimately controlled by leading Tory donor and former party chairman, Lord Ashcroft, has received a £350m contract as part of the government’s COVID-19 vaccination roll-out, openDemocracy has learned.

…The award of a major COVID contract to a firm with close ties to the Tories has sparked further questions about politically connected firms benefiting financially from the UK’s pandemic response.

Last year, a highly critical National Audit Office report found that companies with political links were directed to a “VIP lane” for government contracts where bids were ten times more likely to be successful.

Shadow cabinet office minister Rachel Reeves told openDemocracy: "People are understandably furious seeing businesses owned and run by the friends and donors of the Tory Party being awarded huge multi-million-pound public contracts throughout this pandemic.

“Cronyism, incompetence and waste have been everyday features of this government's approach to outsourcing and ministers show little willingness to learn lessons from the National Audit Office investigations."” For the full article, see here

[10] The low-paid staff of call-center companies Serco and Sitel who responded to recruitment ads calling for a “positive can-do attitude” and a “passion for customer services” have described chaotic training, unwieldy tech systems, and idle shifts for lack of Covid-contact calls to make. Serco’s appointment, in particular, raised eyebrows—both because the company was recently fined more than £1 million ($1.25 million) for failures on another government contract and because junior health minister Edward Argar is a former lobbyist for the company. Originally scheduled to be up and running in June, the track-and-trace system will not now be fully operational until September. In a leaked email, Serco’s chief executive said he wanted the scheme to “cement the position of the private sector” in the NHS supply chain.

Meanwhile, the government’s much-vaunted, custom-built tracing app flunked—after three months of wasted development time and at a cost of several million pounds. It has been abandoned in favor of existing technology from other developers, though it is not clear when this version will be integrated into the government’s track-and-trace program either.

According to the British Medical Journal, by late June local health protection teams had traced nearly eight times more contacts than the outsourced call centers and online services had. Local authority expertise was obvious to anyone who knows anything about public health, Chris Ham told me. The former chief executive of an independent healthcare think tank, the Kings Fund, is now the chair of the Coventry and Warwickshire Health and Care Partnership Board, a nonprofit that coordinates local government and health services in the Midlands. “It is what they are trained to do, bread and butter to them,” he said. “It’s about knowing your community, the sociology, the anthropology of different areas. It cannot be done nationally by people in call centers.” See more in this Guardian article.

[11] See this, for example. 


Ami Clarke

Read more

Related Projects

Risk Related

A series of commissions exploring risk and its social, ecological and economic relations. Read more