Stefaan Verhulst

The Editorial Board at the Financial Times: “The soundtrack of school students marching through Britain’s streets shouting “f*** the algorithm” captured the sense of outrage surrounding the botched awarding of A-level exam grades this year. But the students’ anger towards a disembodied computer algorithm is misplaced. This was a human failure. The algorithm used to “moderate” teacher-assessed grades had no agency and delivered exactly what it was designed to do.

It is politicians and educational officials who are responsible for the government’s latest fiasco and should be the target of students’ criticism….

Sensibly designed, computer algorithms could have been used to moderate teacher assessments in a constructive way. Using past school performance data, they could have highlighted anomalies in the distribution of predicted grades between and within schools. That could have led to a dialogue between Ofqual, the exam regulator, and anomalous schools to come up with more realistic assessments….

There are broader lessons to be drawn from the government’s algo fiasco about the dangers of automated decision-making systems. The inappropriate use of such systems to assess immigration status, policing policies and prison sentencing decisions is a live danger. In the private sector, incomplete and partial data sets can also significantly disadvantage under-represented groups when it comes to hiring decisions and performance measures.

Given the severe erosion of public trust in the government’s use of technology, it might now be advisable to subject all automated decision-making systems to critical scrutiny by independent experts. The Royal Statistical Society and The Alan Turing Institute certainly have the expertise to give a Kitemark of approval or flag concerns.

As ever, technology in itself is neither good nor bad. But it is certainly not neutral. The more we deploy automated decision-making systems, the smarter we must become in considering how best to use them and in scrutinising their outcomes. We often talk about a deficit of trust in our societies. But we should also be aware of the dangers of over-trusting technology. That may be a good essay subject for next year’s philosophy A-level….(More)”.

About: “Landlord Tech—what the real estate industry describes as residential property technology, is leading to new forms of housing injustice. Property technology, or “proptech,” has grown dramatically since 2008, and applies to residential, commercial, and industrial buildings, effectively merging the real estate, technology, and finance industries. By employing digital surveillance, data collection, data accumulation, artificial intelligence, dashboards, and platform real estate in tenant housing and neighborhoods, Landlord Tech increases the power of landlords while disempowering tenants and those seeking shelter.

There are few laws and regulations governing the collection and use of data in the context of Landlord Tech. Because it is generally sold to landlords and property managers, not tenants, Landlord Tech is often installed without notifying or discussing potential harms with tenants and community members. These harms include the possibility that sensitive and personal data can be handed over to the police, ICE, or other law enforcement and government agencies. Landlord Tech can also be used to automate evictions, racial profiling, and tenant harassment. In addition, Landlord Tech is used to abet real estate speculation and gentrification, making buildings more desirable to whiter and wealthier tenants, while feeding real estate and tech companies with property – be that data or real estate. Landlord Tech tracking platforms have increasingly been marketed to landlords as solutions to Covid-19, leading to new forms of residential surveillance….(More)”.

Open access book by Christoph Bartneck, Christoph Lütge, Alan Wagner and Sean Welsh: “This book provides an introduction into the ethics of robots and artificial intelligence. The book was written with university students, policy makers, and professionals in mind but should be accessible for most adults. The book is meant to provide balanced and, at times, conflicting viewpoints as to the benefits and deficits of AI through the lens of ethics. As discussed in the chapters that follow, ethical questions are often not cut and dry. Nations, communities, and individuals may have unique and important perspectives on these topics that should be heard and considered. While the voices that compose this book are our own, we have attempted to represent the views of the broader AI, robotics, and ethics communities.

This book provides an introduction into the ethics of robots and artificial intelligence. The book was written with university students, policy makers, and professionals in mind but should be accessible for most adults. The book is meant to provide balanced and, at times, conflicting viewpoints as to the benefits and deficits of AI through the lens of ethics. As discussed in the chapters that follow, ethical questions are often not cut and dry. Nations, communities, and individuals may have unique and important perspectives on these topics that should be heard and considered. While the voices that compose this book are our own, we have attempted to represent the views of the broader AI, robotics, and ethics communities….(More)”.

Gregory Rosston and Scott J. Wallsten at the Hill: “COVID-19 has, among other things, brought home the costs of the digital divide. Numerous op-eds have offered solutions, including increasing subsidies to schools, providing eligible low-income people with a $50 per month broadband credit, funding more digital literacy classes and putting WiFi on school buses. A House bill would allocate $80 billion to ideas meant to close the digital divide.

The key missing component of nearly every proposal to solve the connectivity problem is evidence — evidence suggesting the ideas are likely to work and ways to use evidence in the future to evaluate whether they did work. Otherwise, we are likely throwing money away. Understanding what works and what doesn’t requires data collection and research now and in the future….

Consider President Trump’s belief in hydroxychloroquine as a cure for the novel coronavirus based simply on his “gut.” That resulted in the government ordering the drug to be produced, distributed to hospitals, and 63 million doses put into a strategic national stockpile.

The well-meaning folks offering up multi-billion dollar broadband plans probably recognize the foolhardiness of the president’s gut-check approach to guiding virus treatment plans. But so far, policy makers and advocates are promoting their own gut beliefs that their proposals will treat the digital divide. An evidence-free approach is likely to cost billions of dollars more and connect fewer people than an evidence-based approach.

It doesn’t have to be this way. The pandemic did not only lay bare the implications of the digital divide, it also created a laboratory for studying how best to bridge the divide. The most immediate problem was how to help kids without home broadband attend distance learning classes. Schools had no time to formally study different options — it was a race to find anything that might help. As a result, schools incidentally ran thousands of concurrent experiments around the country….(More)”.

Essay by Thamy Pogrebinschi: “…In such scenarios, it seems relevant to acknowledge the limits of the state to deal with huge and unpredictable challenges and thus the need to resort to civil society. State capacity cannot be built overnight, but social intelligence is an unlimited and permanently available resource. In recent years, digital technology has multiplied what has been long called social intelligence (Dewey) and is now more often known as collective intelligence (Lévy), the wisdom of crowds (Surowiecki), or democratic reason (Landemore).

Taken together, these concepts point to the most powerful tool available to governments facing hard problems and unprecedented challenges: the sourcing and sharing of knowledge, information, skills, resources, and data from citizens in order to address social and political problems.

The Covid-19 pandemic presents an opportunity to test the potential of social intelligence as fuel for processes of creative collaboration that may aid governments to reinvent themselves and prepare for the challenges that will remain after the virus is gone. By creative collaboration, I mean a range of forms of communication, action, and connection among citizens themselves, between citizens and civil society organizations (CSOs), and between the latter two and their governments, all with the common aim of addressing problems that affect all and that the state for various reasons cannot (satisfactorily) respond to alone.

While several Latin American countries have been stuck in the Covid-19 crisis with governments unable or unwilling to contain it or to reduce its damages, a substantial number of digital democratic innovations have been advanced by civil society in the past few months. These comprise institutions, processes, and mechanisms that rely on digital citizen participation as a means to address social and political problems – and, more recently, also problems of a humanitarian nature….

Between March 16 and July 1 of this year, at least 400 digital democratic innovations were created across 18 countries in Latin America with the specific aim of handling the Covid-19 crisis and mitigating its impact, according to recent data from the LATINNO project. These innovations are essentially mechanisms and processes in which citizens, with the aid of digital tools, are enabled to address social, political, and humanitarian problems related to the pandemic.

Citizens engage in and contribute to three levels of responses, which are based on information, connection, and action. About one-fourth of these digital democratic innovations clearly rely on crowdsourcing social intelligence.

The great majority of those digital innovations have been developed by CSOs. Around 75% of them have no government involvement at all, which is striking in a region known for implementing state-driven citizen participation as a result of the democratization processes that took place in the late 20th century. Civil society has stepped in in most countries, particularly where government responses were absent (Brazil and Nicaragua), slow (Mexico), insufficient due to lack of economic resources (Argentina) or infrastructure (Peru), or simply inefficient (Chile).

Based on these data from 18 Latin American countries, one can observe that digital democratic innovations address challenges posed by the Covid-19 outbreak in five main ways: first, generating verified information and reliable data; second, geolocating problems, needs, and demands; third, mobilizing resources, skills, and knowledge to address those problems, needs, and demands; fourth, connecting demand (individuals and organizations in need) and supply (individuals and organizations willing to provide whatever is needed); and fifth and finally, implementing and monitoring public policies and actions. In some countries, there is a sixth use that cuts across the other five: assisting vulnerable groups such as the elderly, women, children and youth, indigenous peoples, and Afro-descendants….(More)”

Paper by Hamed Khaledi: “This research models governance as a collective intelligence process, particularly as a collective design process. The outcome of this process is a solution to a problem. The solution can be a decision, a policy, a product, a financial plan, etc. The quality (value) of the outcome solution reflects the quality (performance) of the process. Using an analytical model, I identify five mediators (channels) through which, different factors and features can affect the quality of the outcome and thus the process. Based on this model, I propose an asymmetric response surface method that introduces factors to the experimental model considering their plausible effects.

As a proof of concept, I implemented a generic collective design process in a web application and measured the effects of several factors on its performance through online experiments. The results demonstrate the effectiveness of the proposed method. They also show that approval voting is significantly superior to plurality voting. Some studies assert that not the design process, but the designers drive the quality of the outcome. However, this study shows that the characteristics of the design process (e.g. voting schemes) as well as the designers (e.g. expertise and gender) can significantly affect the quality of the outcome. Hence, the outcome quality can be used as an indicator of the performance of the process. This enables us to evaluate and compare governance mechanisms objectively free from fairness criteria….(More)”.

Essay by Yasodara Córdova and Tiago Peixoto: “According to the World Bank’s Digital Dividends report, fewer than 20 percent of digital government projects are successes. Particularly in developing countries, these numbers are often associated with a number of challenges: limited funding, stretched implementation capacity, and political instability, to name a few. Yet, even in developing countries, despite similar conditions, some projects seem to fare better than others. Why is that? 

The projects we have worked with in the global south have followed a similar pattern. While there were successes, many projects have failed. We have learned a few things along the way, that we think relate directly to the success or failure of digital government projects. These are not scientific conclusions, they’re personal impressions based on what we’ve seen and experienced.   

1. Information first, services afterwards

A basic function of digital government is the provision of actionable information concerning public services, by they online or offline (e.g. opening hours, documents required for services, and so on). Even more so in developing countries, where most public services are in-person, paper-based, and often involve multiple steps. Yet, fueled by international rankings and benchmarks, governments are often eager to skip stages in their digital journey. This leads them to attempt, and often fail, to provide transactional digital services, before they can even learn  how to offer basic information about these services. The first step in effective transformation should be offering information to users in a simple and accessible manner. Done well, that forms a good foundation for the next step: delivering digital services.  

2. Prioritise the things that will make the biggest difference

Remember that public service delivery follows a power law distribution: a small number of services account for the vast majority of transactions with government. Which these services are will vary according to country, level of government, and models of public service delivery. When the time comes to decide where to start, don’t rely on cookie-cutter lists of services to be digitized. Instead, find out which ones are the most used, and will have the greatest impact. Start with the ones that can be delivered faster, and that are most likely to make users’ lives easier. 

3. Don’t digitise the mess

The fact that a process exists doesn’t mean it’s a good process. Transformation is an opportunity to radically rethink how things work. We’ve seen examples including, for instance, requiring multiple copies of a single document, or imposing more procedures on women than men to open a business. When there is inefficiency in a service, map the bottlenecks and think about how to streamline the process. Don’t just digitise the bottlenecks, they will keep on being an expensive problem. Resist the temptation to digitise things that should not exist in the first place. …(More)”.

Open access book by Matthias C. Kettemann: “There is order on the internet, but how has this order emerged and what challenges will threaten and shape its future? This study shows how a legitimate order of norms has emerged online, through both national and international legal systems. It establishes the emergence of a normative order of the internet, an order which explains and justifies processes of online rule and regulation. This order integrates norms at three different levels (regional, national, international), of two types (privately and publicly authored), and of different character (from ius cogens to technical standards).

Matthias C. Kettemann assesses their internal coherence, their consonance with other order norms and their consistency with the order’s finality. The normative order of the internet is based on and produces a liquefied system characterized by self-learning normativity. In light of the importance of the socio-communicative online space, this is a book for anyone interested in understanding the contemporary development of the internet….(More)”.

Editorial by Karamjit S. Gill at AI&Society: “Reflecting on the rise of instrumentalism, we learn how it has travelled across the academic boundary to the high-tech culture of Silicon Valley. At its core lies the prediction paradigm. Under the cloak of inevitability of technology, we are being offered the prediction paradigm as the technological dream of public safety, national security, fraud detection, and even disease control and diagnosis. For example, there are offers of facial recognition systems for predicting behaviour of citizens, offers of surveillance drones for ’biometric readings’, ‘Predictive Policing’ is offered as an effective tool to predict and reduce crime rates. A recent critical review of the prediction technology (Coalition for Critical Technology 2020), brings to our notice the discriminatory consequences of predicting “criminality” using biometric and/or criminal legal data.

The review outlines the specific ways crime prediction technology reproduces, naturalizes and amplifies discriminatory outcomes, and why exclusively technical criteria are insufficient for evaluating their risks. We learn that neither predication architectures nor machine learning programs are neutral, they often uncritically inherit, accept and incorporate dominant cultural and belief systems, which are then normalised. For example, “Predictions” based on finding correlations between facial features and criminality are accepted as valid, interpreted as the product of intelligent and “objective” technical assessments. Furthermore, the data from predictive outcomes and recommendations are fed back into the system, thereby reproducing and confirming biased correlations. The consequence of this feedback loop, especially in facial recognition architectures, combined with a belief in “evidence based” diagnosis, is that it leads to ‘widespread mischaracterizations of criminal justice data’ that ‘justifies the exclusion and repression of marginalized populations through the construction of “risky” or “deviant” profiles’…(More).

Essay by Stefania Milan and Emiliano Treré: “Quantification is central to the narration of the COVID-19 pandemic. Numbers determine the existence of the problem and affect our ability to care and contribute to relief efforts. Yet many communities at the margins, including many areas of the Global South, are virtually absent from this number-based narration of the pandemic. This essay builds on critical data studies to warn against the universalization of problems, narratives, and responses to the virus. To this end, it explores two types of data gaps and the corresponding “data poor.” The first gap concerns the data poverty perduring in low-income countries and jeopardizing their ability to adequately respond to the pandemic. The second affects vulnerable populations within a variety of geopolitical and socio-political contexts, whereby data poverty constitutes a dangerous form of invisibility which perpetuates various forms of inequality. But, even during the pandemic, the disempowered manage to create innovative forms of solidarity from below that partially mitigate the negative effects of their invisibility….(More)”.