Berkman Klein Center for Internet and Society: Audio Fishbowl

The pandemic is a portal, the novelist Arundhati Roy wrote in an essay for the Financial Times. “We can choose to walk through it, dragging the carcasses of our prejudice and hatred, our avarice, our data banks and dead ideas, our dead rivers and smoky skies behind us. Or we can walk through lightly, with little luggage, ready to imagine another world. And ready to fight for it.”

In many ways, the coronavirus pandemic has resurfaced and amplified the worst in the world: intensification of surveillance, racism, nationalism, anti-scientism, bigotry.

But something strange has happened as well. Changes, ideas and solutions that were previously deemed impossible have suddenly become possible. Many of these changes still don’t go far enough, come with caveats and fine print, are subject to absurd means-testing, or are only temporary. These aren’t necessarily the changes we want, but they give us a glimpse of what has suddenly become possible.

A number of projects are seeking to capture and document the new possible. In this lunch hour, three of these projects, COVID-19 Policy Response, The New Possible, and Don't Go Back to Normal talk about their experience and debate how we can ensure that the new normal doesn’t turn into the old normal.

To learn more about the Berkman Klein Center, visit https://cyber.harvard.edu

Low-income countries have several systemic disadvantages that cumulatively inhibit their capacity to cope with the spread of COVID-19. These systemic disadvantages, a result of long-term poverty and resource-constrained healthcare systems, are further worsened by other socio-economic outcomes of lockdowns and the spread of infection. BKC hosted a seminar on the economic and healthcare fallouts of COVID-19 in low-income countries, with a specific focus on groups such as women, refugees, and informal laborers, alongside options for international collaboration.

BKC’s Padmashree Gehl Sampath sets the stage and moderates the discussions, joined by BKC’s Yvonne Macpherson, who shares her work on COVID-19’s impact on women and refugees, and Dr. Madani B. Thiam, Chief of Health and Nutrition, UNICEF Myanmar, who speaks on the impact of COVID-19 on health systems capacity and healthcare from his experience in the field.

Ever since Florida 2000, it seems the US cannot hold an election without horror stories about equipment failing, votes lost, and disenfranchised voters says Ben Adida, co-founder and Executive Director of VotingWorks. "Why does a country as powerful and resourceful as the United States have so much trouble running an election?" is a central question to this virtual event. In "Building Better Voting Systems,"

Adida discusses why running an election in the US is particularly challenging. He focuses on voting equipment; it's conventional wisdom that voting machines are universally terrible, and Adida argues that we need to understand how we got here. But it doesn't have to be that way: Adida explains how the US can do much better, and specifically what VotingWorks is working on. Finally, Adida covers how VotingWorks and others are going to help run safe and trustworthy elections in the COVID-19 era.

As the novel coronavirus (COVID-19) sweeps the world in devastating fashion, scientists are scrambling to develop effective vaccines and treatments. But how should those medicines be priced globally?

Following Donald Trump’s “America First” policy with respect to vaccine and drug pricing would be tragic, argue Quentin Palfrey and John Stubbs. Instead, Palfrey and Stubbs propose a pharmaceutical pricing policy modeled on progressive taxation to distribute costs equitably worldwide. This discussion was moderated by Ashveena Gajeelee.

Governments and publics are increasingly asking that tech companies work to address the challenges and adapt to the changes technology has unleashed, from digital security to the impact of the COVID-19 pandemic. At the core of these new expectations is the sense that world-changing technologies must be governed in accordance with a broad ethic of responsibility – to individual users and to society at large.

In this conversation, Jonathan Zittrain was joined by Microsoft President Brad Smith to discuss how big tech might rise to these new challenges and opportunities.

As people turn to news outlets for information, journalists -- and data journalists in particular -- are under pressure to make sense of droves of complicated information.

Data Overload discusses the challenges journalists face obtaining, analyzing, and explaining data about the current pandemic.

Todd Wallack, a Berkman Klein-Nieman Fellow and data journalist at the Boston Globe, is joined by Caroline Chen, who covers health care for ProPublica, and Armand Emamdjomeh, an assignment editor, graphics at the Washington Post.

This event was co-sponsored by the Nieman Foundation.

Fairness and discrimination in algorithmic systems are globally recognized as topics of critical importance. To date, the majority of work in this area starts from an American regulatory perspective defined by the notions of ‘disparate treatment’ and ‘disparate impact.’ But European legal notions of discrimination are not equivalent.

In this talk, Sandra Wachter, Visiting Professor at Harvard Law School and Associate Professor and Senior Research Fellow in Law and Ethics of AI, Big Data, robotics and Internet Regulation at the Oxford Internet Institute (OII) at the University of Oxford, examines EU law and jurisprudence of the European Court of Justice concerning non-discrimination and identifies a critical incompatibility between European notions of discrimination and existing work on algorithmic and automated fairness.

Wachter discusses the evidential requirements for bringing a claim under EU non-discrimination law and propose a statistical test as a baseline to identify and assess potential cases of algorithmic discrimination in Europe.

The identification of bots is an important and complicated task. The bot classifier Botometer was successfully introduced as a way to estimate the number of bots in a given list of accounts and has been frequently used in academic publications.

Given its relevance for academic research, and our understanding of the presence of automated accounts in any given Twitter discourse, Adrian Rauchfleisch and Jonas Kaiser studied Botometer's diagnostic ability over time. To do so, Rauchfleisch and Kaiser collected the Botometer scores for five datasets in two languages (English/German) over three months. For this virtual event, Rauchfleisch and Kaiser discussed their findings and answered questions about the implications of their research.

This virtual talk features Jessica Fjeld, assistant director of the Cyberlaw Clinic and lead author on the “Principled AI” report, in conversation with Ryan Budish, an assistant research director at Berkman Klein and a member of OECD’s AI Governance Expert Group, which proposed high-level AI principles. Fjeld and Budish discuss AI principles both generally (the high-level landscape in which they exist) and in practice (the creation and implementation process for principles.)

“In our data-driven society, it is too easy to assume the transparency of data. Instead, we should approach data sets with an awareness that they are created by humans and their dutiful machines, at a time, in a place, with the instruments at hand, for audiences that are conditioned to receive them,” says Yanni Alexander Loukissas, Assistant Professor of Digital Media in the School of Literature, Media, and Communication at Georgia Tech.

All data are local. The term data set implies something discrete, complete, and portable, but it is none of those things. Examining a series of sources important for understanding public data in the United States—Harvard's Arnold Arboretum, the Digital Public Library of America, UCLA's Television News Archive, and the real estate marketplace Zillow—this talk explains how to analyze data settings rather than data sets.

This talk sets out six principles: all data are local; data have complex attachments to place; data are collected from heterogeneous sources; data and algorithms are inextricably entangled; interfaces recontextualize data; and data are indexes to local knowledge. Then, it provides a set of practical guidelines to follow. These findings are based on a combination of qualitative research on data cultures and exploratory data visualizations. Rebutting the myth of “digital universalism,” this work reminds audiences of the meaning-making power of the local.