Spies Like Us: The Promise and Peril of Crowdsourced Intelligence

Curated on Posted on by Stefaan Verhulst

Book Review by Amy Zegart of “We Are Bellingcat: Global Crime, Online Sleuths, and the Bold Future of News” by Eliot Higgins: “On January 6, throngs of supporters of U.S. President Donald Trump rampaged through the U.S. Capitol in an attempt to derail Congress’s certification of the 2020 presidential election results. The mob threatened lawmakers, destroyed property, and injured more than 100 police officers; five people, including one officer, died in circumstances surrounding the assault. It was the first attack on the Capitol since the War of 1812 and the first violent transfer of presidential power in American history.

Only a handful of the rioters were arrested immediately. Most simply left the Capitol complex and disappeared into the streets of Washington. But they did not get away for long. It turns out that the insurrectionists were fond of taking selfies. Many of them posted photos and videos documenting their role in the assault on Facebook, Instagram, Parler, and other social media platforms. Some even earned money live-streaming the event and chatting with extremist fans on a site called DLive. 

Amateur sleuths immediately took to Twitter, self-organizing to help law enforcement agencies identify and charge the rioters. Their investigation was impromptu, not orchestrated, and open to anyone, not just experts. Participants didn’t need a badge or a security clearance—just an Internet connection….(More)”.

Metroverse

Curated on Posted on by Stefaan Verhulst

About: “Metroverse is an urban economy navigator built at the Growth Lab at Harvard University. It is based on over a decade of research on how economies grow and diversify and offers a detailed look into the specialization patterns of cities.

As a dynamic resource, the tool is continually evolving with new data and features to help answer questions such as:

  • What is the economic composition of my city?
  • How does my city compare to cities around the globe?
  • Which cities look most like mine?
  • What are the technological capabilities that underpin my city’s current economy?
  • Which growth and diversification paths does that suggest for the future?

As city leaders, job seekers, investors and researchers grapple with 21st century urbanization challenges, the answer to these questions are fundamental to understanding the potential of a city.

Metroverse delivers new insights on these questions by placing a city’s technological capabilities and knowhow at the heart of its growth prospects, where the range and nature of existing capabilities strongly influences how future diversification unfolds. Metroverse makes visible what a city is good at today to help understand what it can become tomorrow…(More)”.

To regulate AI, try playing in a sandbox

Curated on Posted on by Stefaan Verhulst

Article by Dan McCarthy: “For an increasing number of regulators, researchers, and tech developers, the word “sandbox” is just as likely to evoke rulemaking and compliance as it is to conjure images of children digging, playing, and building. Which is kinda the point.

That’s thanks to the rise of regulatory sandboxes, which allow organizations to develop and test new technologies in a low-stakes, monitored environment before rolling them out to the general public. 

Supporters, from both the regulatory and the business sides, say sandboxes can strike the right balance of reining in potentially harmful technologies without kneecapping technological progress. They can also help regulators build technological competency and clarify how they’ll enforce laws that apply to tech. And while regulatory sandboxes originated in financial services, there’s growing interest in using them to police artificial intelligence—an urgent task as AI is expanding its reach while remaining largely unregulated. 

Even for all of its promise, experts told us, the approach should be viewed not as a silver bullet for AI regulation, but instead as a potential step in the right direction. 

Rashida Richardson, an AI researcher and visiting scholar at Rutgers Law School, is generally critical of AI regulatory sandboxes, but still said “it’s worth testing out ideas like this, because there is not going to be any universal model to AI regulation, and to figure out the right configuration of policy, you need to see theoretical ideas in practice.” 

But waiting for the theoretical to become concrete will take time. For example, in April, the European Union proposed AI regulation that would establish regulatory sandboxes to help the EU achieve its aim of responsible AI innovation, mentioning the word “sandbox” 38 times, compared to related terms like “impact assessment” (13 mentions) and “audit” (four). But it will likely take years for the EU’s proposal to become law. 

In the US, some well-known AI experts are working on an AI sandbox prototype, but regulators are not yet in the picture. However, the world’s first and (so far) only AI-specific regulatory sandbox did roll out in Norway this March, as a way to help companies comply with AI-specific provisions of the EU’s General Data Protection Regulation (GDPR). The project provides an early window into how the approach can work in practice.

“It’s a place for mutual learning—if you can learn earlier in the [product development] process, that is not only good for your compliance risk, but it’s really great for building a great product,” according to Erlend Andreas Gjære, CEO and cofounder of Secure Practice, an information security (“infosec”) startup that is one of four participants in Norway’s new AI regulatory sandbox….(More)”

Scientific publishing’s new weapon for the next crisis: the rapid correction

Curated on Posted on by Stefaan Verhulst

Gideon Meyerowitz-Katz and James Heathers at STATNews: “If evidence of errors does emerge, the process for correcting or withdrawing a paper tends to be alarmingly long. Late last year, for example, David Cox, the IBM director of the MIT-IBM Watson AI Lab, discovered that his name was included as an author on two papers he had never written. After he wrote to the journals involved, it took almost three months for them to remove his name and the papers themselves. In cases of large-scale research fraud, correction times can be measured in years.

Imagine now that the issue with a manuscript is not a simple matter of retracting a fraudulent paper, but a more complex methodological or statistical problem that undercuts the study’s conclusions. In this context, requests for clarification — or retraction — can languish for years. The process can outlast both the tenure of the responsible editor, resetting the clock on the entire ordeal, or the journal itself can cease publication, leaving an erroneous article in the public domain without oversight, forever….

This situation must change, and change quickly. Any crisis that requires scientific information in a hurry will produce hurried science, and hurried science often includes miscalculated analyses, poor experimental design, inappropriate statistical models, impossible numbers, or even fraud. Having the agility to produce and publicize work like this without having the ability to correct it just as quickly is a curiously persistent oversight in the global scientific enterprise. If corrections occur only long after the research has already been used to treat people across the world, what use are they at all?

There are some small steps in the right direction. The open-source website PubPeer aggregates formal scientific criticism, and when shoddy research makes it into the literature, hordes of critics may leave comments and questions on the site within hours. Twitter, likewise, is often abuzz with spectacular scientific critiques almost as soon as studies go up online.

But these volunteer efforts are not enough. Even when errors are glaring and obvious, the median response from academic journals is to deal with them grudgingly or not at all. Academia in general takes a faintly disapproving tone of crowd-sourced error correction, ignoring the fact that it is often the only mechanism that exists to do this vital work.

Scientific publishing needs to stop treating error-checking as a slightly inconvenient side note and make it a core part of academic research. In a perfect world, entire departmental sections would be dedicated to making sure that published research is correct and reliable. But even a few positions would be a fine start. Young researchers could be given kudos not just for every citation in their Google scholar profile but also for every post-publication review they undertake….(More)”

When Graphs Are a Matter of Life and Death

Curated on Posted on by Stefaan Verhulst

Essay by  Hannah Fry at the NewYorker: “John Carter has only an hour to decide. The most important auto race of the season is looming; it will be broadcast live on national television and could bring major prize money. If his team wins, it will get a sponsorship deal and a chance to start making some real profits for a change.

There’s just one problem. In seven of the past twenty-four races, the engine in the Carter Racing car has blown out. An engine failure live on TV will jeopardize sponsorships—and the driver’s life. But withdrawing has consequences, too. The wasted entry fee means finishing the season in debt, and the team won’t be happy about the missed opportunity for glory. As Burns’s First Law of Racing says, “Nobody ever won a race sitting in the pits.”

One of the engine mechanics has a hunch about what’s causing the blowouts. He thinks that the engine’s head gasket might be breaking in cooler weather. To help Carter decide what to do, a graph is devised that shows the conditions during each of the blowouts: the outdoor temperature at the time of the race plotted against the number of breaks in the head gasket. The dots are scattered into a sort of crooked smile across a range of temperatures from about fifty-five degrees to seventy-five degrees.

When Graphs Are a Matter of Life and Death

The upcoming race is forecast to be especially cold, just forty degrees, well below anything the cars have experienced before. So: race or withdraw?

This case study, based on real data, and devised by a pair of clever business professors, has been shown to students around the world for more than three decades. Most groups presented with the Carter Racing story look at the scattered dots on the graph and decide that the relationship between temperature and engine failure is inconclusive. Almost everyone chooses to race. Almost no one looks at that chart and asks to see the seventeen missing data points—the data from those races which did not end in engine failure.

Image may contain Plot

As soon as those points are added, however, the terrible risk of a cold race becomes clear. Every race in which the engine behaved properly was conducted when the temperature was higher than sixty-five degrees; every single attempt that occurred in temperatures at or below sixty-five degrees resulted in engine failure. Tomorrow’s race would almost certainly end in catastrophe.

One more twist: the points on the graph are real but have nothing to do with auto racing. The first graph contains data compiled the evening before the disastrous launch of the space shuttle Challenger, in 1986….(More)”.

Examining the Intersection of Behavioral Science and Advocacy

Curated on Posted on by Stefaan Verhulst

Introduction to Special Collection of the Behavioral Scientist by Cintia Hinojosa and Evan Nesterak: “Over the past year, everyone’s lives have been touched by issues that intersect science and advocacy—the pandemic, climate change, police violence, voting, protests, the list goes on. 

These issues compel us, as a society and individuals, toward understanding. We collect new data, design experiments, test our theories. They also inspire us to examine our personal beliefs and values, our roles and responsibilities as individuals within society. 

Perhaps no one feels these forces more than social and behavioral scientists. As members of fields dedicated to the study of social and behavioral phenomena, they are in the unique position of understanding these issues from a scientific perspective, while also navigating their inevitable personal impact. This dynamic brings up questions about the role of scientists in a changing world. To what extent should they engage in advocacy or activism on social and political issues? Should they be impartial investigators, active advocates, something in between? 

t also raises other questions, like does taking a public stance on an issue affect scientific integrity? How should scientists interact with those setting policies? What happens when the lines between an evidence-based stance and a political position become blurred? What should scientists do when science itself becomes a partisan issue? 

To learn more about how social and behavioral scientists are navigating this terrain, we put out a call inviting them to share their ideas, observations, personal reflections, and the questions they’re grappling with. We gave them 100-250 words to share what was on their mind. Not easy for such a complex and consequential topic.

The responses, collected and curated below, revealed a number of themes, which we’ve organized into two parts….(More)”.

Sandwich Strategy

Curated on Posted on by Stefaan Verhulst

Article by the Accountability Research Center: “The “sandwich strategy” describes an interactive process in which reformers in government encourage citizen action from below, driving virtuous circles of mutual empowerment between pro-accountability actors in both state and society.

The sandwich strategy relies on mutually-reinforcing interaction between pro-reform actors in both state and society, not just initiatives from one or the other arena. The hypothesis is that when reformers in government tangibly reduce the risks/costs of collective action, that process can bolster state-society pro-reform coalitions that collaborate for change. While this process makes intuitive sense, it can follow diverse pathways and encounter many roadblocks. The dynamics, strengths and limitations of sandwich strategies have not been documented and analyzed systematically. The figure below shows a possible pathway of convergence and conflict between actors for and against change in both state and society….(More)”.

sandwich strategy

Selected Readings on the Use of Artificial Intelligence in the Public Sector

Curated on Posted on by Uma

By Kateryna Gazaryan and Uma Kalkar

The Living Library’s Selected Readings series seeks to build a knowledge base on innovative approaches for improving the effectiveness and legitimacy of governance. This curated and annotated collection of recommended works focuses on algorithms and artificial intelligence in the public sector.

As Artificial Intelligence becomes more developed, governments have turned to it to improve the speed and quality of public sector service delivery, among other objectives. Below, we provide a selection of recent literature that examines how the public sector has adopted AI to serve constituents and solve public problems. While the use of AI in governments can cut down costs and administrative work, these technologies are often early in development and difficult for organizations to understand and control with potential harmful effects as a result. As such, this selected reading explores not only the use of artificial intelligence in governance but also its benefits, and its consequences.

Readings are listed in alphabetical order.

Berryhill, Jamie, Kévin Kok Heang, Rob Clogher, and Keegan McBride. “Hello, World: Artificial intelligence and its use in the public sector.” OECD Working Papers on Public Governance no. 36 (2019): https://doi.org/10.1787/726fd39d-en.

This working paper emphasizes the importance of defining AI for the public sector and outlining use cases of AI within governments. It provides a map of 50 countries that have implemented or set in motion the development of AI strategies and highlights where and how these initiatives are cross-cutting, innovative, and dynamic. Additionally, the piece provides policy recommendations governments should consider when exploring public AI strategies to adopt holistic and humanistic approaches.

Kuziemski, Maciej, and Gianluca Misuraca. “AI Governance in the Public Sector: Three Tales from the Frontiers of Automated Decision-Making in Democratic Settings.” Telecommunications Policy 44, no. 6 (2020): 101976. 

Kuziemski and Misuraca explore how the use of artificial intelligence in the public sector can exacerbate existing power imbalances between the public and the government. They consider the European Union’s artificial intelligence “governance and regulatory frameworks” and compare these policies with those of Canada, Finland, and Poland. Drawing on previous scholarship, the authors outline the goals, drivers, barriers, and risks of incorporating artificial intelligence into public services and assess existing regulations against these factors. Ultimately, they find that the “current AI policy debate is heavily skewed towards voluntary standards and self-governance” while minimizing the influence of power dynamics between governments and constituents. 

Misuraca, Gianluca, and Colin van Noordt. “AI Watch, Artificial Intelligence in Public Services: Overview of the Use and Impact of AI in Public Services in the EU.” 30255 (2020).

This study provides “evidence-based scientific support” for the European Commission as it navigates AI regulation via an overview of ways in which European Union member-states use AI to enhance their public sector operations. While AI has the potential to positively disrupt existing policies and functionalities, this report finds gaps in how AI gets applied by governments. It suggests the need for further research centered on the humanistic, ethical, and social ramification of AI use and a rigorous risk assessment from a “public-value perspective” when implementing AI technologies. Additionally, efforts must be made to empower all European countries to adopt responsible and coherent AI policies and techniques.

Saldanha, Douglas Morgan Fullin, and Marcela Barbosa da Silva. “Transparency and Accountability of Government Algorithms: The Case of the Brazilian Electronic Voting System.” Cadernos EBAPE.BR 18 (2020): 697–712.

Saldanha and da Silva note that open data and open government revolutions have increased citizen demand for algorithmic transparency. Algorithms are increasingly used by governments to speed up processes and reduce costs, but their black-box  systems and lack of explanability allows them to insert implicit and explicit bias and discrimination into their calculations. The authors conduct a qualitative study of the “practices and characteristics of the transparency and accountability” in the Brazilian e-voting system across seven dimensions: consciousness; access and reparations; accountability; explanation; data origin, privacy and justice; auditing; and validation, precision and tests. They find the Brazilian e-voting system fulfilled the need to inform citizens about the benefits and consequences of data collection and algorithm use but severely lacked in demonstrating accountability and opening algorithm processes for citizen oversight. They put forth policy recommendations to increase the e-voting system’s accountability to Brazilians and strengthen auditing and oversight processes to reduce the current distrust in the system.

Sharma, Gagan Deep, Anshita Yadav, and Ritika Chopra. “Artificial intelligence and effective governance: A review, critique and research agenda.” Sustainable Futures 2 (2020): 100004.

This paper conducts a systematic review of the literature of how AI is used across different branches of government, specifically, healthcare, information, communication, and technology, environment, transportation, policy making, and economic sectors. Across the 74 papers surveyed, the authors find a gap in the research on selecting and implementing AI technologies, as well as their monitoring and evaluation. They call on future research to assess the impact of AI pre- and post-adoption in governance, along with the risks and challenges associated with the technology.

Tallerås, Kim, Terje Colbjørnsen, Knut Oterholm, and Håkon Larsen. “Cultural Policies, Social Missions, Algorithms and Discretion: What Should Public Service Institutions Recommend?” Part of the Lecture Notes in Computer Science book series (2020).

Tallerås et al. examine how the use of algorithms by public services, such as public radio and libraries, influence broader society and culture. For instance, to modernize their offerings, Norway’s broadcasting corporation (NRK) has adopted online platforms similar to popular private streaming services. However, NRK’s filtering process has faced “exposure diversity” problems that narrow recommendations to already popular entertainment and move Norway’s cultural offerings towards a singularity. As a public institution, NRK is required to “fulfill […] some cultural policy goals,” raising the question of how public media services can remain relevant in the era of algorithms fed by “individualized digital culture.” Efforts are currently underway to employ recommendation systems that balance cultural diversity with personalized content relevance that engage individuals and uphold the socio-cultural mission of public media.

Vogl, Thomas, Seidelin Cathrine, Bharath Ganesh, and Jonathan Bright. “Smart Technology and the Emergence of Algorithmic Bureaucracy: Artificial Intelligence in UK Local Authorities.” Public administration review 80, no. 6 (2020): 946–961.

Local governments are using “smart technologies” to create more efficient and effective public service delivery. These tools are twofold: not only do they help the public interact with local authorities, they also streamline the tasks of government officials. To better understand the digitization of local government, the authors conducted surveys, desk research, and in-depth interviews with stakeholders from local British governments to understand reasoning, processes, and experiences within a changing government framework. Vogl et al. found an increase in “algorithmic bureaucracy” at the local level to reduce administrative tasks for government employees, generate feedback loops, and use data to enhance services. While the shift toward digital local government demonstrates initiatives to utilize emerging technology for public good, further research is required to determine which demographics are not involved in the design and implementation of smart technology services and how to identify and include these audiences.

Wirtz, Bernd W., Jan C. Weyerer, and Carolin Geyer. “Artificial intelligence and the public sector—Applications and challenges.” International Journal of Public Administration 42, no. 7 (2019): 596-615.

The authors provide an extensive review of the existing literature on AI uses and challenges in the public sector to identify the gaps in current applications. The developing nature of AI in public service has led to differing definitions of what constitutes AI and what are the risks and benefits it poses to the public. As well, the authors note the lack of focus on the downfalls of AI in governance, with studies tending to primarily focus on the positive aspects of the technology. From this qualitative analysis, the researchers highlight ten AI applications: knowledge management, process automation, virtual agents, predictive analytics and data visualization, identity analytics, autonomous systems, recommendation systems, digital assistants, speech analytics, and threat intelligence. As well, they note four challenge dimensions—technology implementation, laws and regulation, ethics, and society. From these applications and risks, Wirtz et al. provide a “checklist for public managers” to make informed decisions on how to integrate AI into their operations. 

Wirtz, Bernd W., Jan C. Weyerer, and Benjamin J. Sturm. “The dark sides of artificial intelligence: An integrated AI governance framework for public administration.” International Journal of Public Administration 43, no. 9 (2020): 818-829.

As AI is increasingly popularized and picked up by governments, Wirtz et al. highlight the lack of research on the challenges and risks—specifically, privacy and security—associated with implementing AI systems in the public sector. After assessing existing literature and uncovering gaps in the main governance frameworks, the authors outline the three areas of challenges of public AI: law and regulations, society, and ethics. Last, they propose an “integrated AI governance framework” that takes into account the risks of AI for a more holistic “big picture” approach to AI in the public sector.

Zuiderwijk, Anneke, Yu-Che Chen, and Fadi Salem. “Implications of the use of artificial intelligence in public governance: A systematic literature review and a research agenda.” Government Information Quarterly (2021): 101577.

Following a literature review on the risks and possibilities of AI in the public sector, Zuiderwijk, Chen, and Salem design a research agenda centered around the “implications of the use of AI for public governance.” The authors provide eight process recommendations, including: avoiding superficial buzzwords in research; conducting domain- and locality-specific research on AI in governance; shifting from qualitative analysis to diverse research methods; applying private sector “practice-driven research” to public sector study; furthering quantitative research on AI use by governments; creating “explanatory research designs”; sharing data for broader study; and adopting multidisciplinary reference theories. Further, they note the need for scholarship to delve into best practices, risk management, stakeholder communication, multisector use, and impact assessments of AI in the public sector to help decision-makers make informed decisions on the introduction, implementation, and oversight of AI in the public sector.

Digitalization as a common good. Contribution to an inclusive recovery

Curated on Posted on by Stefaan Verhulst

Essay by Julia Pomares, Andrés Ortega & María Belén Abdala: “…The pandemic has accelerated the urgency of a new social contract for this era at national, regional, and global levels, and such a pact clearly requires a digital dimension. The Spanish government, for example, proposes that by 2025, 100 megabits per second should be achieved for 100% of the population. A company like Telefónica, for its part, proposes a “Digital Deal to build back better our societies and economies” to achieve a “fair and inclusive digital transition,” both for Spain and Latin America.

The pandemic and the way of coping with and overcoming it has also emphasized and aggravated the significance of different types of digital and connectivity gaps and divides, between countries and regions of the world, between rural and urban areas, between social groups, including income and gender-related gaps, and between companies (large and small), which need to be addressed and bridged in these new social digital contracts. For the combination of digital divides and the pandemic amplify social disparities and inequalities in various spheres of life. Digitalization can contribute to enlarge those divides, but also to overcome them.

Common good

In 2016, the UN, through its Human Rights Council and General Assembly, qualified access to the internet as a basic fundamental human right, from which all human rights can also be defended. In 2021, the Italian Presidency of the G20 has set universal access to the internet as a goal of the group.

We use the concept of common good, in a non-legal but economic sense, following Nobel Laureate Elinor Ostrom 6 who refers to the nature of use and not of ownership. In line with Ostrom, digitalization and connectivity as a common good respond to three characteristics:

  • It is non-rivalrous: Its consumption by anyone does not reduce the amount available to others (which in digitalization and connectivity is true to a certain extent, since it also relies on huge but limited storage and processing centers, and also on network capacity, both in the access and backbone network. It is the definition of service, where a distinction has to be made between the content of what is transmitted, and the medium used.)
  • It is non-excludable: It is almost impossible to prevent anyone from consuming it.
  • It is available, more or less, all over the world….(More)”.

How Low and Middle-Income Countries Are Innovating to Combat Covid

Curated on Posted on by Stefaan Verhulst

Article by Ben Ramalingam, Benjamin Kumpf, Rahul Malhotra and Merrick Schaefer: “Since the Covid-19 pandemic hit, innovators around the world have developed thousands of novel solutions and practical approaches to this unprecedented global health challenge. About one-fifth of those innovations have come from low- and middle-income countries across sub-Saharan Africa, South Asia, and Latin America, according to our analysis of WHO data, and they work to address the needs of poor, marginalized, or excluded communities at the so-called bottom of the pyramid.

Over the past year we’ve been able to learn from and support some of those inspiring innovators. Their approaches are diverse in scope and scale and cover a vast range of pandemic response needs — from infection prevention and control to community engagement, contract tracing, social protection, business continuity, and more.

Here we share seven lessons from those innovators that offer promising insights not only for the ongoing Covid response but also for how we think about, manage, and enable innovation.

1. Ensure that your solutions are sensitive to social and cultural dynamics. 

Successful innovations are relevant to the lived realities of the people they’re intended to help. Socially and culturally sensitive design approaches see greater uptake and use. This is true in both resource-constrained and resource-rich environments.

Take contact tracing in Kenya. In a context where more than half of all residents use public transportation every day, the provider of a ticketing app for Nairobi’s bus fleets adapted its software to collect real-time passenger data. The app has been used across one of the world’s most mobile populations to trace Covid-19 cases, identify future clusters, trigger automated warnings to exposed passengers, and monitor the maximum number of people that could safely be allowed in each vehicle….(More)”.