By Sebastian Bornschlegl
Let us take over! How Jack Stilgoe wants to put the public back into the driving seat of innovation
Book Review: Stilgoe, J. (2020). Who’s Driving Innovation? New Technologies and the Collaborative State. Cham: Palgrave Macmillan.
We are accustomed to a story of a mobility revolution told by companies like Uber, Waymo, and Tesla year after year: Autonomous vehicles (AVs) are just around the corner, probably available this very year. But something always seems to get in the way of these great promises. In 2020 the global COVID-19 pandemic and accompanying regulations like social distancing and lockdowns brought the development and testing of AVs to a sudden halt (Ohnsman, 2020a; Wiggers, 2020). Tesla’s CEO Elon Musk made headlines multiple times, defying the lockdown of a factory in Fremont and anti-COVID measures in general (Boudette, 2020; Boudette & Flitter, 2020; Lambert, 2020). Serious health risks for workers were accepted in the name of innovation. AV technology is presented as too disruptive, too important, too essential for society than to be stopped by regulations. This story of inevitable technological progress against all odds is aptly challenged by Stilgoe (2020) in his pamphlet Who’s Driving Innovation? New Technologies and the Collaborative State.
Stilgoe is one of the leading STS scholars working on automated driving and artificial intelligence (AI). His new book builds on his previous work on self-driving cars (2018, 2019) in which he highlighted the technology’s infrastructural interdependence and social complexity. In his pamphlet the author explores how emergent technologies are governed and links current trends to our sociotechnical past. He employs AVs and AI as prime examples while focussing mostly on the US and UK context. Stilgoe poses an essential question: How can societies hold innovators responsible for developing technologies that benefit all of society instead of a small elite?
The situation today appears like the exact opposite: Private companies drive and control the process of innovation in the high-tech sector. They either circumscribe regulations or enroll the public sector for private interests. Stilgoe challenges the techno-deterministic visions of society enacted by companies like Google and Facebook across five chapters. His critique is focused on the way private entities organize innovation, which often compromises social equality and welfare in favor of profit and autonomy. But Stilgoe does not stop at problematization, he also explores potential alternatives for democratizing innovation. The concept of the Collaborative State is a tentative, normative program for the governance of emergent technologies that puts governments and the public back into the driving seat of innovation.
An extraordinary fatal accident which occurred in 2018 acts as the entry point to the issue of current technological innovations: The killing of Elaine Herzberg by a self-driving car operated by Uber in Tempe, Arizona. Blame was assigned to everybody, but the decision to test an apparently dangerous technology without sufficient safeguards was not questioned. This “tombstone mentality” (Stilgoe, 2020, p. 4) ignores the need for regulations until it’s too late. The emerging theme of companies not taking responsibility for their innovations while promising big benefits for society is present throughout all chapters. Stilgoe argues that “[i]n the absence of any outside involvement, science and technology will tend to reinforce rather than close inequalities.” (Stilgoe, 2020, p. 32) For example Uber’s investment in self-driving taxis opens up the possibility of getting rid of its underpaid, precarious drivers altogether. In a similar manner AI might be employed for decision making in a variety of social contexts not because it is more just but simply because it is cheaper than human labour.
His dire prospect is connected to a set of issues with the current governance of innovation. New technologies are presented as quick fixes for specialized problems and get assessed according to their possible risks. If something goes wrong, technology’s shortcomings are frequently labeled as unintended side effects instead of its inherent risks. But as Stilgoe argues, many bugs are indeed a deliberate feature serving the interests of private innovators: Facebook is designed around privacy infringements and the free productivity services provided by Google feed into its advertising business. These risks might be unknown to consumers and regulators alike. The author links the opaqueness of powerful software companies to “economies of scale” (Stilgoe, 2020, p. 29), meaning that current high-tech innovations like AI and machine learning depend on vast centralization efforts by private companies. These entities position themselves as the solution to the need for data collection and computational power. But once high-tech companies get hold of public data, they enforce full control over it – often without any public oversight.
While AV innovators like Musk claim that they will use their power to develop revolutionary self-driving software, Stilgoe points out that their promises have not yet manifested at all. The killing of Herzberg exemplifies that AVs cannot perform even mundane tasks like breaking on time. Thought experiments like the “Trolley problem”, where the machine is faced with a dilemma situation and has to decide on whom to run over, “provide a convenient distraction from a real debate about the limits of technologies and the responsibility of engineers.” (Stilgoe, 2020, p. 45). Colleagues like JafariNaimi (2018) also criticize the utilitarian framing of treating lives as calculative variable, resulting in AI programmed to kill. Stilgoe’s argument culminates in the realization that “the dream of instant autonomy promises to change to world without changing the world.” (Stilgoe, 2020, p. 47) Innovations like machine learning would have to be combined with processes of social learning and infrastructural developments in order to really benefit society. If cars shall drive automatically, our built environments as well as ways of living will need to change radically – and not just cars.
The last chapter on the concept of the Collaborative State positions policymaking and public experimentation as a way of democratizing innovation. The author tries to translate insights from how the NHS governs innovation in the medical sector to how governments could regulate disruptive innovations in general. This would result in “public policy as a form of grand experiment” (Stilgoe, 2020, p. 58). Stilgoe is convinced that governments and policymakers should lead the way of innovation in order to make sure that the public actually profits from private innovators. Public participation as well as flexible, pro-active policies are the tools he proposes for this end.
Is there a chance that AI and AVs will actually revolutionize societies across the globe? The author reformulates this question: Can we put effective regulations into place so that these innovations get democratized, mitigating inequalities instead of reinforcing them?
The critique is on point, but “how” remains an open question
Sticking to the formal characteristics of a pamphlet, the book presents a normative program supported by fitting empirical evidence and pointed argumentation instead of an in-depth analyses or case study. Stilgoe ties together past work on the governance of innovation and the politics of technology like Winner’s (1980) famous example of the tomato harvester in order to understand the societal impact of current bleeding edge technology. Just like the tomato harvesting machine radically changed the plants, plantations and connected human labour, emergent technologies will not simply replace existing ones but alter the fabric of society in significant ways. This puts Stilgoe’s book in relation to STS work on engineering cultures like Hughe’s (1987) study on large technological systems as well as Law’s (1987) concept of heterogenous engineering. Stilgoe is convinced that technology is always embedded into the larger context of society, thus being dependent on built environments, social relations as well as infrastructures. He highlights that engineering is not only a technical process, but also a social one. To paraphrase Latour’s (1988) paradigm of science being politics by other means, Stilgoe presents the invention and governance of technology as a way of doing politics.
The innovative character of the book lies in updating past debates for the 2020s and focussing on AI and self-driving cars as one of the most discussed technologies today. Even though these innovations are presented as disruptive and revolutionizing, Stilgoe shows that technologies and their corresponding politics are never inevitable. He formulates a critique of technological determinism that remains comprehensible considering the shortness of the book. His warning is as relevant as ever: The testing of AVs is already taken up again (Ohnsman, 2020b) and some companies claim that these vehicles could be the solution to the COVID-19 mobility crisis (ITU News, 2020) by reducing human contact in the transportation of essential goods. One can imagine the imminent tensions between private innovators and regulations in times of crisis.
Stilgoe’s pamphlet is a call to democratize innovations in order to strengthen public participation and welfare. This message is consistently picked up in all of the chapters. The Collaborative State as means to this end remains vague though, which in part is owed to the short form. For me a main concern regarding this concept is the belief that experimentation and public participation lead to good governance by the state per se. For once, the lack of regulations might be a deliberate decision in order to cater to the interests of private stakeholders. The ways certain governments handle the current crisis should be a telling warning that politics often do not have public welfare in mind. Secondly participation experiments come with their own set of issues (Bogner, 2012), namely that representing the public under lab conditions tends to reinforce experts’ hypotheses instead of yielding novel insights. Stilgoe critically reflects that “experiments in public are also experiments on the public” (2020, p. 59), thus encompassing profound ethical challenges. It remains to be seen how collaborative forms of governing technology can be organized effectively and whether civil society interacts out of its own interests instead of being called by expert bodies.
Sebastian Bornschlegl BA BA is currently studying the STS program at the University of Vienna coming from a humanities background. He is preparing his MA thesis on semi-autonomous buses and their users. In his spare time he is podcasting about politics and culture for Schirmchen & Streusel.
Bogner, A. (2012). The Paradox of Participation Experiments. Science, Technology, & Human Values, 37(5), 506–527. https://doi.org/10.1177/0162243911430398
Boudette, N. E. (2020, May 8). Tesla Tells Workers It Will Reopen California Factory Despite County Order. Retrieved August 11, 2020, from The New York Times website: https://www.nytimes.com/2020/05/08/business/economy/tesla-coronavirus-factory-alameda.html
Boudette, N. E., & Flitter, E. (2020, May 9). Elon Musk Lashes Out at Officials Keeping Tesla Plant Closed Over Virus. Retrieved August 11, 2020, from The New York Times website: https://www.nytimes.com/2020/05/09/business/coronavirus-elon-musk-tesla-california.html
Hughes, T. P. (1987). The Evolution of Large Technological Systems. In W. E. Bijker, T. P. Hughes, & T. Pinch (Eds.), The social construction of technological systems. New directions in the sociology and history of technology (pp. 51–82). Cambridge/London: MIT Press.
ITU News. (2020, April 29). COVID-19: Where are the self-driving cars and trucks? Retrieved August 11, 2020, from ITU News website: https://news.itu.int/covid-19-where-are-the-self-driving-cars-and-trucks/
JafariNaimi, N. (2018). Our Bodies in the Trolley’s Path, or Why Self-driving Cars Must *Not* Be Programmed to Kill. Science, Technology, & Human Values, 43(2), 302–323. https://doi.org/10.1177/0162243917718942
Lambert, F. (2020, July 14). Tesla sees spike in COVID-19 “exposure,” data leak shows over 130 cases and more “affected.” Retrieved August 11, 2020, from Electrek website: https://electrek.co/2020/07/14/tesla-spike-covid-19-cases-data-leak/
Latour, B. (1988). The Pasteurization of France. Cambridge: Harvard University Press.
Law, J. (1987). Technology and heterogeneous engineering. The case of Portuguese expansion. In W. E. Bijker, T. P. Hughes, & T. Pinch (Eds.), The social construction of technological systems. New directions in the sociology and history of technology (pp. 111–134). Cambridge/London: MIT Press.
Ohnsman, A. (2020a, March 17). Waymo Halts Phoenix Robotic-Ride Service With Backup Human Drivers Due To Coronavirus – But Fully Driverless Vans Keep Rolling. Retrieved August 11, 2020, from Forbes website: https://www.forbes.com/sites/alanohnsman/2020/03/17/waymo-halts-phoenix-robotic-ride-service-with-backup-human-drivers-due-to-covid-19-but-keeps-fully-driverless-vans-rolling/
Ohnsman, A. (2020b, May 7). Waymo’s Self-Driving Vans Ready To Roll After COVID-19 Shutdown – But No Robotaxi Rides. Retrieved August 11, 2020, from Forbes website: https://www.forbes.com/sites/alanohnsman/2020/05/07/waymos-self-driving-vans-to-roll-again-after-covid-19-shutdownbut-no-robotaxi-rides/
Stilgoe, J. (2018). Machine learning, social learning and the governance of self-driving cars. Social Studies of Science, 48(1), 25–56. https://doi.org/10.1177/0306312717741687
Stilgoe, J. (2019). Self-driving cars will take a while to get right. Nature Machine Intelligence, 1(5), 202–203. https://doi.org/10.1038/s42256-019-0046-z
Stilgoe, J. (2020). Who’s Driving Innovation? New Technologies and the Collaborative State. Cham: Palgrave Macmillan. https://doi.org/10.1007/978-3-030-32320-2
Wiggers, K. (2020, April 28). The challenges of developing autonomous vehicles during a pandemic. Retrieved August 21, 2020, from VentureBeat website: https://venturebeat.com/2020/04/28/challenges-of-developing-autonomous-vehicles-during-coronavirus-covid-19-pandemic/
Winner, L. (1980). Do Artifacts Have Politics? Daedalus, 109(1), 121–136.