Security is often conflated with secrecy, cutting communities and democratic processes out of technology design and oversight. This note outlines the academic research that shows why secrecy leads to less secure technologies and disparate impacts. Co-authored with Julia Slupska of Oxford Internet Institute, Jeanette Lowrie of Tech Workers Coalition, and Professor Deian Stefan of UCSD’s Crypto and Security Group.
Cedric Whitney and I published a report last week, “Broken Promises of Civic Innovation: Technological, Organizational, Fiscal, and Equity Challenges of GE Current CityIQ” as part of the work we do at the Institute for Practical Ethics.
The claimed public benefits of the Intelligent Cities project, billed as the ‘the World’s Largest Smart City Platform’ by the city, were promoted as creating data for sustainability, promoting civic innovation, and saving energy on lighting. The City states that this project ‘is a tremendous technological benefit to the city and our citizens’, and that ‘from easier parking and decreased traffic congestion, enhanced public safety and environmental monitoring, enhanced bicycle route planning, to enhanced urban and real estate development planning, this platform can improve the quality of life in our city and boost economic growth’.
The city has already spent three years and millions of dollars on this platform implementation. Yet these aspirations for civic empowerment and sustainability data have not been realized. This project has been limited by technical breakdowns, organizational limitations, and an opportunity structure that adversely affects lower-income San Diegans.
Instead, the city is left with a surveillance system that pervasively records video in public thoroughfares and near homes, workplaces, and places of worship – and the city, not citizens, access and use the data. Ongoing data recording incurs costs of data storage, data transmission, and the electricity required to maintain operations of the networked computer system.
This report summarizes the results of investigation of the system at the Institute of Practical Ethics and Design Lab at UC San Diego into the implications of CityIQ smart streetlights for privacy and inequality.
- The planning data produced by the streetlights system is highly unreliable three years and approximately $7 million in loan repayments into the implementation. Smart streetlights, in effect, are only reliable as a video and audio surveillance system.
- Three of the data modalities that are vital to the purported use cases (pedestrian, parking, and traffic data) have major operational flaws, including < 0.5% (only 12) of the cameras currently reporting pedestrian data.
- Despite promises to support civic innovation, technical infrastructures and organizational support are lacking and require ongoing and expanded financial investments.
- The City states that the platform has not led to a single running, externally created application on the platform. It is not in a position to be used by citizens or entrepreneurs.
- City of San Diego is using tax payer funds to beta test for GE Current, bearing risks and costs of a largely untested and highly complex public infrastructure. It is also paying taxpayer funds as GE Current learns how to fix its system for future clients and deployments.
San Diego is the first city to deploy the CityIQ smart city platform. The problems in implementation, including technical flaws in the CityIQ platform, can be attributed to problems that arise in the development and stabilization of any new complex technological system. However, policy makers and tax payers should evaluate whether the costs of unstable, early stage systems should be borne by the company that markets and profit from the system. High-tech companies often market their internet platforms as “permanently beta” and subject to adaptation, but costs and risks of system failure are currently borne by the City of San Diego.
 Smart Streetlights Program – City of San Diego. https://www.sandiego.gov/sustainability/energy-and-water-efficiency/programs-projects/smart-city.
 This research was conducted by Lilly Irani, Cedric Whitney, Simrandeep Singh, Elizabeth Quepones, Lauran Irion, and Steven Rick at UC San Diego.
 The City has stated that CityIQ audio recording is available but not been enabled.
The Committee for the Anthropology of Science, Technology, and Computing (CASTAC) of the General Anthropology Division (GAD) and The Society for the Anthropology of Work (SAW), announce Lilly Irani as the winner of the 2019 Diana Forsythe Prize for her book, Chasing Innovation: Making Entrepreneurial Citizens in Modern India (Princeton University Press 2019).
From the award committee citation:
“Chasing Innovation is a fearlessly ambitious work of scholarship that weaves together history, ethnography, and critique of a seductive vision of entrepreneurial citizenship. Through its pages, Lilly Irani illustrates how discourses of innovation were articulated with the distinctive context of a liberalized India hungry to climb global chains of value. Marked, at times, by a raw and searching reflexivity as its author reflects on the failures of imagination produced by her own socialization as a tech worker, the book pushes back against the “subsumption of hope” by innovation and points to mass politics—for all its inefficiencies—as the true locus of democratic futures.
Irani is a careful ethnographer who gets inside the optimistic dreams of entrepreneurs, whose impetus to “move fast and break things” in their speculative world-making betrays a certain innocence about the violence of the market economy. Her work brings us into a high-end design studio where the free will of innovators relies on unfree labors of devalued service staff and on the extraction of solutions from subaltern subjects who are framed as improvisers rather than as innovators. The key actors in Chasing Innovation “attempt to stabilize, manage, and profit from uncertainties and futurities” even as they themselves are constrained by the design thinking through which innovation for the developing world is increasingly channeled.”
CS Soong, host of Against the Grain, hosted me for a one-hour conversation on innovation, democracy, autocracy, and development. The show is called “Innovation and its Discontents” and it broadcast Aug 20. You can listen here or through your favorite podcast app.
Dorothy Howard and I co-authored a paper on the politics of qualitative research as a practice of extracting knowledge or practicing solidarity. The paper won an honorable mention at CHI 2019.
Abstract: This paper investigates a hidden dimension of research with real world stakes: research subjects who care — sometimes deeply — about the topic of the research in which they participate. They manifest this care, we show, by managing how they are represented in the research process, by exercising politics in shaping knowledge production, and sometimes in experiencing trauma in the process. We draw first-hand reflections on participation in diversity research on Wikipedia, transforming participants from objects of study to active negotiators of research process. We depict how care, vulnerability, harm, and emotions shape ethnographic and qualitative data. We argue that, especially in reflexive cultures, research subjects are active agents with agendas, accountabilities, and political projects of their own. We propose ethics of care and collaboration to open up new possibilities for knowledge production and socio-technical intervention in HCI.
My book Chasing Innovation is out with Princeton University Press. You can also access the penultimate proof as a PDF over at escholarship. The book will be released in a South Asia edition distributed through Penguin Random House (date TBA).
Chris Kelty, author of Two Bits: The Cultural Significance of Open Source, interviewed me about the book over at the CAMP (Communication, Media and Performance) Anthropology blog. We talk about collective forms in tech production and how they are tied to the state, capital, development, and utopia.
STS scholars Lee Vinsel and Patrick McCrary organized a workshop to explore and critically examining the idea of the 4th Industrial Revolution. As part of a collection of essays coming out of the workshop, I’ve published “Hype, Profit, Labor, and Agency” .
“The hype of inevitable AI and total automation immobilizes AMT workers from demanding better wages or improvement to their work conditions. Like the threat of workers made available cheaply in other parts of the world, public assumptions about AI and automation immobilize grassroots, democratic participation in shaping the future of work. This hype and the broader narratives of novelty and disjuncture that often accompanies it can prematurely disarm existing ethical and justice frameworks by which citizens can make claims about technological futures.”
New Scientist invited me to comment on the activism of Google workers around military contracts, sexual harassment in the work place, and contractor rights. I argue for coalitions between tech workers and those communities already surviving what Safiya Noble calls “algorithms of oppression.” See the short piece here
Wired writer Miranda Katz explains how Amazon Mechanical Turk workers seek new cooperative forms of production and collective action, in collaboration with Niloufar Salehi (Stanford, CS), me, and Micheal Bernstein (Stanford, CS)
Tariq Ghani and Katherine Sacco of UCI Anthropology interviewed me for the American Anthropological Association’s podcast Anthropod. On the 25 minute interview podcast: “Lilly Irani discusses the human labor behind artificial intelligence technology. Irani helped create a platform called Turkopticon to support workers on Amazon Mechanical Turk, a website that outsources micro data processing work. Irani also talks about her current book project on entrepreneurialism and national development in India.”
In 2017, I spoke at the conference of the UC Collaboratory for Ethnographic Design on Turkopticon and the politics of writing labor as an ethnographer. Ghani and Sacco interviewed as part of a 3 part series on the intersections between ethnography and design.