Coronavirus: how the pandemic has exposed AI’s limitations

Curated on Posted on by Stefaan Verhulst

Kathy Peach at The Conversation: “It should have been artificial intelligence’s moment in the sun. With billions of dollars of investment in recent years, AI has been touted as a solution to every conceivable problem. So when the COVID-19 pandemic arrived, a multitude of AI models were immediately put to work.

Some hunted for new compounds that could be used to develop a vaccine, or attempted to improve diagnosis. Some tracked the evolution of the disease, or generated predictions for patient outcomes. Some modelled the number of cases expected given different policy choices, or tracked similarities and differences between regions.

The results, to date, have been largely disappointing. Very few of these projects have had any operational impact – hardly living up to the hype or the billions in investment. At the same time, the pandemic highlighted the fragility of many AI models. From entertainment recommendation systems to fraud detection and inventory management – the crisis has seen AI systems go awry as they struggled to adapt to sudden collective shifts in behaviour.

The unlikely hero

The unlikely hero emerging from the ashes of this pandemic is instead the crowd. Crowds of scientists around the world sharing data and insights faster than ever before. Crowds of local makers manufacturing PPE for hospitals failed by supply chains. Crowds of ordinary people organising through mutual aid groups to look after each other.

COVID-19 has reminded us of just how quickly humans can adapt existing knowledge, skills and behaviours to entirely new situations – something that highly-specialised AI systems just can’t do. At least yet….

In one of the experiments, researchers from the Istituto di Scienze e Tecnologie della Cognizione in Rome studied the use of an AI system designed to reduce social biases in collective decision-making. The AI, which held back information from the group members on what others thought early on, encouraged participants to spend more time evaluating the options by themselves.

The system succeeded in reducing the tendency of people to “follow the herd” by failing to hear diverse or minority views, or challenge assumptions – all of which are criticisms that have been levelled at the British government’s scientific advisory committees throughout the pandemic…(More)”.

How to engage with policy makers: A guide for academics in the arts and humanities

Curated on Posted on by Stefaan Verhulst

Institute for Government: “The Arts and Humanities Research Council and the Institute for Government have been working in partnership for six years on the Engaging with Government programme – a three-day course for researchers in the arts and humanities. This programme helps academics develop the knowledge and skills they need to engage effectively with government and parliamentary bodies at all levels, along with the other organisations involved in the policy-making process. We, in turn, have learned a huge amount from our participants, who now form an active alumni network brimming with expertise about how to engage with policy in practice. This guide brings together some of that learning.

Arts and humanities researchers tend to have fewer formal and established routes into government than scientists. But they can, and do, engage productively in policy making. They contribute both expertise (advice based on knowledge of a field) and evidence (facts and information) and provide new ways of framing policy debates that draw on philosophical, cultural or historical perspectives.

As this guide shows, there are steps that academics can take to improve their engagement with public policy and to make it meaningful for their research. While these activities may involve an investment of time, they offer the opportunity to make a tangible difference, and are often a source of great satisfaction and inspiration for further work.

The first part of this guide describes the landscape of policy making in the UK and some of the common ways academics can engage with it.

Part two sets out six lessons from the Engaging with Government programme, illustrated with practical examples from our alumni and speaker network. These lessons are:

  • Understand the full range of individuals and groups involved in policy making – who are the key players and who do they talk to?
  • Be aware of the political context – how does your research fit in with current thinking on the issue?
  • Communicate in ways that policy makers find useful – consider your audience and be prepared to make practical recommendations.
  • Develop and maintain networks – seek to make connections with people who share your policy interest, both in person and online.
  • Remember that you are the expert – be prepared to share your general knowledge of a subject as well as your specific research.
  • Adopt a long-term perspective – you will need to be open-minded and patient to engage successfully….(More)”.

Do nudgers need budging? A comparative analysis of European smart meter implementation

Curated on Posted on by Stefaan Verhulst

Paper by Sarah Giest: “Nudging is seen to complement or replace existing policy tools by altering people’s choice architectures towards behaviors that align with government aims, but has fallen short in meeting those targets. Crucially, governments do not nudge citizens directly, but need private agents to nudge their consumers. Based on this notion, the paper takes on an institutional approach towards nudging. Rather than looking at the relationship between nudger and nudgee, the research analyses the regulatory and market structures that affect nudge implementation by private actors, captured by the ‘budge’ idea. Focusing on the European energy policy domain, the paper analyses the contextual factors of green nudges that are initiated by Member States, and implemented by energy companies. The findings show that in the smart meter context, there are regulatory measures that affect implementation of smart meters and that government has a central role to ‘budge’, due to the dependence on private agents….(More)”.

Hammer or nudge? New brief on international policy options for COVID-19

Curated on Posted on by Stefaan Verhulst

Paper by Luc Soete: “…But over time the scientific comments given on TV and radio in my two home countries, the Netherlands and Belgium, as well as neighbouring Germany and France, became dominated by each country’s own, national virology and epidemiological experts explaining how their country’s approach to ‘flattening the curve’ and bringing down the reproduction rate was best, it became clear, even to a non-expert in the field like myself, that many of the science-based policies used to contain COVID-19 were first and foremost based on ‘hypotheses’. With the exception of Germany, not really on facts. And as Anthony Fauci, Director of the US National Institute of Allergy and Infectious Disease, probably the world’s most respected virologist once put it: “Data is real. The model is hypothesis.”

So at the risk of being an ultracrepidarian – an old word which has suddenly risen in popularity – it seemed appropriate to have a closer, more critical look at the science-based policy advice during this COVID-19 pandemic. For virologists and epidemiologists, the logical approach to a new, unknown but highly infectious virus such as SARS-CoV-2, spreading globally at pandemic speed, is ‘the hammer’: the tool to crush down quickly and radically through extreme measures (social distancing, confinement, lockdown, travel restrictions) the spread of the virus and get the transmission rate’s value as far as possible below. The stricter the confinement measures, the better.

For a social scientist and social science-based policy adviser, a hammer represents anything but a useful tool to approach society or the economy with. Her or his preference will rather go to measures, such as ‘nudges’ which alter people’s behaviour in a predictable way without coercion. Actually, the first COVID-19 measure was based on a typical ‘nudge’: improving hand hygiene among healthcare workers which was now enlarged to the whole population. ‘Nudging’ in the face of a new virus such as SARS-CoV-2 will consist of making sure incremental policy measures build up to a societal behavioural change, starting from hand hygiene, social distancing, to confinement and various forms of lockdown. It will be crucial to measure the additional, marginal impact of each measure in its contribution to the overall reduction in the transmission of the virus. Introducing all measures at once, as in the case of the ‘hammer’ strategy, subsequently provides little useful information on the effectiveness of each measure ( on the contrary, in fact). In a period of deconfinement, one now has little information on which measures are likely to be the most effective. From a nudge perspective, achieving a change in social behaviour with respect to physical distancing: the so-called one-and-a-half metre society, will be an essential variable and measuring its impact on the spreading of the virus crucial. One of the reasons is that full adoption of such physical distancing automatically and without the need of coercion, will prevent the occurrence of large or smaller social gatherings without authorities having to specify the rules. This is implicit in the principle of nudging: it will be the providers, the entrepreneurs of personal service sectors who will have to come up with organisational innovations enabling physical distancing in the safe delivery of such services.

Most noteworthy, however, is the purely national setting within which most virology and epidemiological science-based policy advice is currently framed. This contrasts sharply with the actual scientific research in the field which is today purely global, based on shared data and open access. For years now, epidemiological studies have taken individual countries as ‘containers’ for data collection and data analysis. It is also the national setting that provides the framework for estimating the capacity of medical facilities, especially the total number of available intensive care units needed to handle COVID-19 patients in each country. In the case of Europe and as a result, it has led to the reintroduction of internal borders which had ‘disappeared’ 25 years ago for fear of cross-border contamination. Doing so, COVID-19 has undermined the notion of European values. This policy brief is my attempt to clarify the situation….(More)”.

Mapping Mobility Functional Areas (MFA) using Mobile Positioning Data to Inform COVID-19 Policies

Curated on Posted on by Stefaan Verhulst

EU Science Hub: “This work introduces the concept of data-driven Mobility Functional Areas (MFAs) as geographic zones with a high degree of intra-mobility exchanges. Such information, calculated at European regional scale thanks to mobile data, can be useful to inform targeted re-escalation policy responses in cases of future COVID-19 outbreaks (avoiding large-area or even national lockdowns). In such events, the geographic distribution of MFAs would define territorial areas to which lockdown interventions could be limited, with the result of minimizing socio-economic consequences of such policies. The analysis of the time evolution of MFAs can also be thought of as a measure of how human mobility changes not only in intensity but also in patterns, providing innovative insights into the impact of mobility containment measures. This work presents a first analysis for 15 European countries (14 EU Member States and Norway)….(More)”.

Ethical and societal implications of algorithms, data, and artificial intelligence: a roadmap for research

Curated on Posted on by Stefaan Verhulst

Report by the Nuffield Foundation and the Leverhulme Centre for the Future of Intelligence:” The aim of this report is to offer a broad roadmap for work on the ethical and societal implications of algorithms, data, and AI (ADA) in the coming years. It is aimed at those involved in planning, funding, and pursuing research and policy work related to these technologies. We use the term ‘ADA-based technologies’ to capture a broad range of ethically and societally relevant technologies based on algorithms, data, and AI, recognising that these three concepts are not totally separable from one another and will often overlap. A shared set of key concepts and concerns is emerging, with widespread agreement on some of the core issues (such as bias) and values (such as fairness) that an ethics of algorithms, data, and AI should focus on. Over the last two years, these have begun to be codified in various codes and sets of ‘principles’. Agreeing on these issues, values and high-level principles is an important step for ensuring that ADA-based technologies are developed and used for the benefit of society. However, we see three main gaps in this existing work: (i) a lack of clarity or consensus around the meaning of central ethical concepts and how they apply in specific situations; (ii) insufficient attention given to tensions between ideals and values; (iii) insufficient evidence on both (a) key technological capabilities and impacts, and (b) the perspectives of different publics.”….(More)”.

Data is Dangerous: Comparing the Risks that the United States, Canada and Germany See in Data Troves

Curated on Posted on by Stefaan Verhulst

Paper by Susan Ariel Aaronson: “Data and national security have a complex relationship. Data is essential to national defense — to understanding and countering adversaries. Data underpins many modern military tools from drones to artificial intelligence. Moreover, to protect their citizens, governments collect lots of data about their constituents. Those same datasets are vulnerable to theft, hacking, and misuse. In 2013, the Department of Defense’s research arm (DARPA) funded a study examining if “ the availability of data provide a determined adversary with the tools necessary to inflict nation-state level damage. The results were not made public. Given the risks to the data of their citizens, defense officials should be vociferous advocates for interoperable data protection rules.

This policy brief uses case studies to show that inadequate governance of personal data can also undermine national security. The case studies represent diverse internet sectors affecting netizens differently. I do not address malware or disinformation, which are also issues of data governance, but have already been widely researched by other scholars. I illuminate how policymakers, technologists, and the public are/were unprepared for how inadequate governance spillovers affected national security. I then makes some specific recommendations about what we can do about this problem….(More)”.

The European data market

Curated on Posted on by Stefaan Verhulst

European Commission: “It was the first European Data Market study (SMART 2013/0063) contracted by the European Commission in 2013 that made a first attempt to provide facts and figures on the size and trends of the EU data economy by developing a European data market monitoring tool.

The final report of the updated European Data Market (EDM) study (SMART 2016/0063) now presents in detail the results of the final round of measurement of the updated European Data Market Monitoring Tool contracted for the 2017-2020 period.

Designed along a modular structure, as a first pillar of the study, the European Data Market Monitoring Tool is built around a core set of quantitative indicators to provide a series of assessments of the emerging market of data at present, i.e. for the years 2018 through 2020, and with projections to 2025.

The key areas covered by the indicators measured in this report are:

  • The data professionals and the balance between demand and supply of data skills;
  • The data companies and their revenues;
  • The data user companies and their spending for data technologies;
  • The market of digital products and services (“Data market”);
  • The data economy and its impacts on the European economy.
  • Forecast scenarios of all the indicators, based on alternative market trajectories.

Additionally, as a second major work stream, the study also presents a series of descriptive stories providing a complementary view to the one offered by the Monitoring Tool (for example, “How Big Data is driving AI” or “The Secondary Use of Health Data and Data-driven Innovation in the European Healthcare Industry”), adding fresh, real-life information around the quantitative indicators. By focusing on specific issues and aspects of the data market, the stories offer an initial, indicative “catalogue” of good practices of what is happening in the data economy today in Europe and what is likely to affect the development of the EU data economy in the medium term.

Finally, as a third work stream of the study, a landscaping exercise on the EU data ecosystem was carried out together with some community building activities to bring stakeholders together from all segments of the data value chain. The map containing the results of the landscaping of the EU data economy as well as reports from the webinars organised by the study are available on the www.datalandscape.eu website….(More)”.

Public understanding and perceptions of data practices: a review of existing research

Curated on Posted on by Stefaan Verhulst

Report by Helen Kennedy, Susan Oman, Mark Taylor, Jo Bates & Robin Steedman: “The ubiquitous collection and use of digital data is said to have wide-ranging effects. As these practices expand, interest in how the public perceives them has begun to grow. Understanding public views of data
practices is considered to be important, to ensure that data works ‘for people and society’ (the mission of the Ada Lovelace Institute) and is ‘a force for good’ (an aim of the government Centre for Data Ethics and
Innovation)

To improve understanding of public views of data practices, we conducted a review of original empirical research into public perceptions of, attitudes toward and feelings about data practices. We use the term ‘data practices’ to refer to the systematic collection, analysis and sharing of data and the outcomes of these processes. The data at the centre of such practices is often personal data, and related research often focuses on this data. Our review also covered related phenomena such as AI and facial recognition.

We carried out a systematic search of online academic research databases and a manual search, that began with literature with which we were already familiar, and then snowballed out. Our review covered a broad
range of academic disciplines and grey literature – that is, literature produced by independent, civil society, third sector, governmental or commercial organisations or by academics for non-academic audiences. It focused on the past five years. We excluded a) literature about children’s understandings and perceptions of data practices because this is a specialist area beyond our remit, and b) literature focused on the health domain because high quality syntheses of literature focusing on this domain already exist. The grey literature we reviewed focused on the UK, whereas academic literature was international….(More)”.

Are Citizens’ Assemblies the Answer to the Climate Crisis?

Curated on Posted on by Stefaan Verhulst

Judy Dempsey’s Strategic Europe: “Mathilde Bouyé associate at the Climate Program Of The World Resources Institute: “…the impact of citizens’ deliberation depends on the link to decisionmaking, which varies with each country’s democratic culture. The UK climate assembly informed powerful parliamentary committees, while the French government created a precedent by committing to send the Citizens’ Convention on Climate’s proposals for adoption “without any filter….”

Jan Eichhorn,  Research Director Of D|Part and Senior Lecturer in Social Policy at The University Of Edinburgh: “The climate crisis is so complex that no single action can be the answer to it. However, because of the complexity, formats that can connect otherwise distant actors meaningfully can play a very helpful role. Citizens’ assemblies fit that bill.

If well designed, such assemblies connect expertise with life realities, broaden the horizon of policymakers on what publics may be willing or even excited to consider, and enable publics to learn about options they did not know about. Rather than stoking divisions between people and businesses or between activists and state officials, they can foster common ground and create shared purpose, which is needed to combat comprehensive challenges like the climate crisis….”

Tim Hughes, Director of Involve: “…they are only one way in which people can be—and need to be—involved in decisionmaking. Underpinning citizens’ assemblies are the principles of participation—people being involved in the decisions that affect their lives—and deliberation—people sharing and testing ideas through inclusive and respectful conversations.

It is these principles that we need to build into decisionmaking at all levels of society in order to develop the ideas, energy, and ownership to answer the crisis.”

Mariann Őry,  Head Of The Foreign Desk And Senior Editor At Magyar Hírlap: “Citizens’ initiatives have proven to be effective in reaching a number of goals, but the pressure they can put on stakeholders is not always enough.

It’s not even the most reliable political force: remember that the enthusiasm and momentum of the climate protests has basically vanished since the start of the coronavirus crisis, as if people simply lost interest—though this is surely not the case. A difference can be made on the level of political leaders and, very importantly, on the level of the biggest actors of industry….(More)”.