Authors:
Eleanor Dare
One word from this pandemic will, I predict, haunt us above all others. The horrific Covid infection and death rates in the U.S. and U.K., the poverty and inequality magnified by the virus, the violence of state indecisiveness and equivocation and the inadequacy and inertia of neoliberal governments in the wealthiest nations to address systemic deprivation and disease, converge upon the hideous portmanteau Covidpreneur.
Covidpreneur is a new word "formed by the combination of entrepreneur and Covid-19, used to describe individuals or businesses that thrive and innovate in a pandemic environment" [1]. Covidpreneur is a word I cannot shake from my head, one that aligns with a ruthless minority of self-serving "innovators" and opportunists—often the keenest proponents of data dashboards and AI-driven "edtech," whom I have often heard referring to the pandemic as an "opportunity." The purported opportunity is usually for self-advancement or furthering personal profit, yet ineptly results in the opposite scenario—redundancies in higher education and many other industries; the algorithmic U.K. school-exam-result fiasco, in which automated results mechanisms were found to replicate discrimination; track-and-trace profiteering with very little actual tracking and tracing; PPE supply failures supplied to us by get-rich-quick multimillionaires; vaccine shortages; and oscillating school and university Covid-19 policies, among so many other systemic failings. Far from being an opportunity, what we have encountered over the past 13 months has been likened to two alternate realities, "with some worried about their next haircut and others worried about their next meal" [2]. Here we can define inequality, especially that relating to health, as "systematic, avoidable and unjust differences in health and well-being between different groups of people," arising "because of the conditions in which we are born, grow, live, work and become older, which influence our opportunities for good health and how we think, feel and act" [2].
The spread of the virus has been exacerbated by a housing crisis that has been with us in the U.K. since Prime Minister Thatcher first privatized state housing via the Housing Act of August 1980, failing to replace housing lost to "right to buy" while also decimating and selling off public services and basic infrastructure. My horror, as someone who might once have defined themselves as a creative technologist (but is increasingly loath to do so, such is the distaste for STEAM's right-wing associations), is that Covidpreneurism is inextricably entangled with STEM and STEAM agendas, and with the failure of academia to distance itself from the neoliberal teleology of corporate America. Government initiatives to compete with other nations—in particular, China—such as the America COMPETES Act of 2007 and, in the U.K., Prime Minister Theresa May's stated commitment to an AI revolution in which she promised to put AI at the center, "creating up to 2,500 AI and data master's courses, enabling the workforce to up-skill and 'contribute to the ongoing AI revolution'" [3].
Covidpreneurism is inextricably entangled with STEM and STEAM agendas.
This is in the context of the AI equivalent of a Cold War—era space race (and with Brexit, we now have the added element of direct competition with France and Germany in the tech wars). For U.K. Prime Minister Theresa May, back in June 2019, technology was the primary agent of change:
It will shape the world to come and will solve some of today's most pressing issues. Air pollution and congestion, quality healthcare, security and equal access to jobs are just a few examples of critical concerns we see in our cities and communities across the UK [3].
"Let's come together," May announced, "to create a tech nation that truly is worlds apart" [3]. Of course, there was no insight into the many critiques of AI (or, to be more accurate, machine learning) as a set of statistical processes that reinforce discrimination. The U.K. is certainly worlds apart from those countries that acted decisively to save lives, such as New Zealand and Iceland. May's uncritical technological determinism and that of her successor, Prime Minister Boris Johnson, is driven by the ideology that brought us the abject failure of track and trace. Sadly, what makes the U.K. stand apart, and so shamefully close to Trump/s failed response to the pandemic, is the inability of our technological commitments—to AI and big data, smart cities and tech hubs, accelerators and hackathons, industrial strategies and STEM agendas—to make any difference at all to the impact of the Covid pandemic beyond launching a handful of Covidpreneurs to wealth beyond imagining. All the while, child poverty, housing, healthcare, education, and employment face ongoing crises. An added irony is that the "disruptive" innovation of social media appears to be a significant factor in the growth of anti-science conspiracy theories and ideologies, which will slow the uptake of the one great contribution to slowing the pandemic: vaccines. With Trump's presidency now behind us, let's hope we can now reevaluate neoliberal techno-determinism, and with it, recognize that the countries that protected their populations most effectively are characterized by a commitment to social justice and equality, not Covidpreneurial opportunism or AI cold wars with China.
Since the pandemic began in the U.K. in early 2020, I have been collaborating with Alexandra Antonopoulou to write about and visualize our experiences of lockdown, as well as our memories of pre-pandemic life. We have used a chatbot I developed with the Python Chatterbot library as a collaborator in this project, as well as a range of algorithms to convert speech into images and images into captions. The images I have included here are part of my ongoing research into the limits of symbolic representation, and by now, well-known issues of machine learning systems replicating inequality and the status quo.
For the image in Figure 1 (please forgive the low quality, which is part of an algorithmic visualization process), I used the sentence "I ran and ran as fast and as long as I could, pumping my arms and straining at the neck" to generate it.
 |
Figure 1. AI-generated text-to-image representation of the words: "I ran and ran as fast and as long as I could, pumping my arms and straining at the neck." |
On January 13, 2021, I then used another algorithm to caption the image in Figure 1, and found, on successive attempts, that the algorithm assumed the presence of male agents, often characterized as skateboarders, tennis players, and golfers—a replication of what seems like Silicon Valley culture, not my memories of running to meet Alexandra at Hilly Fields in South East London (Figure 2).
 |
Figure 2. AI-generated evaluation of agents and objects in the image. |
I then entered the following words into the text-to-image algorithm: "Machine learning representations always seem to frame our collaboration as active men—golfers, tennis players, skateboarders; grafting a Silicon Valley lifestyle onto us, seeing us/not seeing us through its own filters," which resulted in the image shown in Figure 3.
 |
Figure 3. AI-generated text-to-image representation of the words: "Machine learning representations always seem to frame our collaboration as active men - golfers, tennis players, skateboarders; grafting a Silicon Valley lifestyle onto us, seeing us/ not seeing us through its own filters." |
Much of my work has investigated AI through practice, but after 16 years of such activity, I am more ambivalent than ever about deploying such algorithms. Their entanglement with the reproduction of the status quo, and with racism and misogyny, is now well known. A wide range of brilliant writers, such as Ruha Benjamin [4,5] and Safiya Umoja Noble [6], have elucidated that association. The value of this strategy—using tools that reduce us all to data points and commodities—is increasingly unclear, as time and time again AI either underwhelms or offends. As someone who once described themselves as a creative technologist, the future of my own practice is rightly uncertain.
References
1. internews.org. 'Covidpreneur'; https://internews.org/covid-19/glossary/covidpreneur
2. Campos-Matos, I., Newton, J., and Doyle, Y. An opportunity to address inequalities: learning from the first months of the COVID-19 pandemic. Public Health Matters. Oct. 29, 2020; https://publichealthmatters.blog.gov.uk/2020/10/29/an-opportunity-to-address-inequalities-learning-from-the-first-months-of-the-covid-19-pandemic/
3. Poole, G. and Roughan, A. May's latest commitments to tech are welcome, but the next PM has more work to do. Tech Newstatesman. Jun. 10, 2019; https://tech.newstatesman.com/business/theresa-may-uk-tech 10th June, 2019
4. Benjamin, R. Captivating Technology: Race, Carceral Technoscience, and Liberatory Imagination in Everyday Life. Duke Univ. Press, Durham:London, 2019.
5. Benjamin, R. Race After Technology, Abolitionist Tools for the New Jim Code, Polity Press, Cambridge, U.K., 2019.
6. Noble, S.U. Algorithms of Oppression. NYU Press, 2018.
Author
Eleanor Dare has taught computer programming and themes relating to digital theory at a range of U.K.-based colleges, including the Royal College of Art, Goldsmiths, the University of Derby, the Open University, UCL, London College of Communication, and the Creative Computing Institute. Her current writing and programming research are concerned with the limits of symbolic logic, representation, and creative nonfiction, as well as the implications of virtual reality. [email protected]
Copyright held by author
The Digital Library is published by the Association for Computing Machinery. Copyright © 2021 ACM, Inc.
Post Comment
No Comments Found