Reflecting the Past, Shaping the Future: Making AI Work for International Development

Curated on Posted on by Stefaan Verhulst

USAID Report: “We are in the midst of an unprecedented surge of interest in machine learning (ML) and artificial intelligence (AI) technologies. These tools, which allow computers to make data-derived predictions and automate decisions, have become part of daily life for billions of people. Ubiquitous digital services such as interactive maps, tailored advertisements, and voice-activated personal assistants are likely only the beginning. Some AI advocates even claim that AI’s impact will be as profound as “electricity or fire” that it will revolutionize nearly every field of human activity. This enthusiasm has reached international development as well. Emerging ML/AI applications promise to reshape healthcare, agriculture, and democracy in the developing world. ML and AI show tremendous potential for helping to achieve sustainable development objectives globally. They can improve efficiency by automating labor-intensive tasks, or offer new insights by finding patterns in large, complex datasets. A recent report suggests that AI advances could double economic growth rates and increase labor productivity 40% by 2035. At the same time, the very nature of these tools — their ability to codify and reproduce patterns they detect — introduces significant concerns alongside promise.

In developed countries, ML tools have sometimes been found to automate racial profiling, to foster surveillance, and to perpetuate racial stereotypes. Algorithms may be used, either intentionally or unintentionally, in ways that result in disparate or unfair outcomes between minority and majority populations. Complex models can make it difficult to establish accountability or seek redress when models make mistakes. These shortcomings are not restricted to developed countries. They can manifest in any setting, especially in places with histories of ethnic conflict or inequality. As the development community adopts tools enabled by ML and AI, we need a cleareyed understanding of how to ensure their application is effective, inclusive, and fair. This requires knowing when ML and AI offer a suitable solution to the challenge at hand. It also requires appreciating that these technologies can do harm — and committing to addressing and mitigating these harms.

ML and AI applications may sometimes seem like science fiction, and the technical intricacies of ML and AI can be off-putting for those who haven’t been formally trained in the field. However, there is a critical role for development actors to play as we begin to lean on these tools more and more in our work. Even without technical training in ML, development professionals have the ability — and the responsibility — to meaningfully influence how these technologies impact people.

You don’t need to be an ML or AI expert to shape the development and use of these tools. All of us can learn to ask the hard questions that will keep solutions working for, and not against, the development challenges we care about. Development practitioners already have deep expertise in their respective sectors or regions. They bring necessary experience in engaging local stakeholders, working with complex social systems, and identifying structural inequities that undermine inclusive progress. Unless this expert perspective informs the construction and adoption of ML/AI technologies, ML and AI will fail to reach their transformative potential in development.

This document aims to inform and empower those who may have limited technical experience as they navigate an emerging ML/AI landscape in developing countries. Donors, implementers, and other development partners should expect to come away with a basic grasp of common ML techniques and the problems ML is uniquely well-suited to solve. We will also explore some of the ways in which ML/AI may fail or be ill-suited for deployment in developing-country contexts. Awareness of these risks, and acknowledgement of our role in perpetuating or minimizing them, will help us work together to protect against harmful outcomes and ensure that AI and ML are contributing to a fair, equitable, and empowering future…(More)”.

Following Fenno: Learning from Senate Candidates in the Age of Social Media and Party Polarization

Curated on Posted on by Stefaan Verhulst

David C.W. Parker  at The Forum: “Nearly 40 years ago, Richard Fenno published Home Style, a seminal volume explaining how members of Congress think about and engage in the process of representation. To accomplish his task, he observed members of Congress as they crafted and communicated their representational styles to the folks back home in their districts. The book, and Fenno’s ensuing research agenda, served as a clarion call to move beyond sophisticated quantitative analyses of roll call voting and elite interviews in Washington, D.C. to comprehend congressional representation. Instead, Fenno argued, political scientists are better served by going home with members of Congress where “their perceptions of their constituencies are shaped, sharpened, or altered” (Fenno 1978, p. xiii). These perceptions of constituencies fundamentally shape what members of Congress do at home and in Washington. If members of Congress are single-minded seekers of reelection, as we often assume, then political scientists must begin with the constituent relationship essential to winning reelection. Go home, Fenno says, to understand Congress.

There are many ways constituency relationships can be understood and uncovered; the preferred method for Fenno is participant observation, which he variously terms as “soaking and poking” or “just hanging around.” Although it sounds easy enough to sit and watch, good participant observation requires many considerations (as Fenno details in a thorough appendix to Home Style). In this appendix, and in another series of essays, Fenno grapples forthrightly with the tough choices researchers must consider when watching and learning from politicians.

In this essay, I respond to Fenno’s thought-provoking methodological treatise in Home Style and the ensuing collection of musings he published as Watching Politicians: Essays on Participant Observation. I do so for three reasons: First, I wish to reinforce Fenno’s call to action. As the study of political science has matured, it has moved away from engaging with politicians in the field across the various sub-fields, favoring statistical analyses. “Everyone cites Fenno, but no one does Fenno,” I recently opined, echoing another scholar commenting on Fenno’s work (Fenno 2013, p. 2; Parker 2015, p. 246). Unfortunately, that sentiment is supported by data (Grimmer 2013, pp. 13–19; Curry 2017). Although quantitative and formal analyses have led to important insights into the study of political behavior and institutions, politics is as important to our discipline as science. And in politics, the motives and concerns of people are important to witness, not just because they add complexity and richness to our stories, but because they aid in theory generation.1 Fenno’s study was exploratory, but is full of key theoretical insights relevant to explaining how members of Congress understand their constituencies and the ensuing political choices they make.

Second, to “do” participant observation requires understanding the choices the methodology imposes. This necessitates that those who practice this method of discovery document and share their experiences (Lin 2000). The more the prospective participant observer can understand the size of the choice set she faces and the potential consequences at each decision point in advance, the better her odds of avoiding unanticipated consequences with both immediate and long-term research ramifications. I hope that adding my cumulative experiences to this ongoing methodological conversation will assist in minimizing both unexpected and undesirable consequences for those who follow into the field. Fenno is open about his own choices, and the difficult decisions he faced as a participant observer. Encouraging scholars to engage in participant observation is only half the battle. The other half is to encourage interested scholars to think about those same choices and methodological considerations, while acknowledging that context precludes a one-size fits all approach. Fenno’s choices may not be your choices – and that might be just fine depending upon your circumstances. Fenno would wholeheartedly agree.

Finally, Congress and American politics have changed considerably from when Fenno embarked on his research in Home Style. At the end of his introduction, Fenno writes that “this book is about the early to mid-1970s only. These years were characterized by the steady decline of strong national party attachments and strong local party organizations. … Had these conditions been different, House members might have behaved differently in their constituencies” (xv). Developments since Fenno put down his pen include political parties polarizing to an almost unprecedented degree, partisan attachments strengthening among voters, and technology emerging to change fundamentally how politicians engage with constituents. In light of this evolution of political culture in Washington and at home, it is worth considering the consequences for the participant-observation research approach. Many have asked me if it is still possible to do such work in the current political environment, and if so, what are the challenges facing political scientists going into the field? This essay provides some answers.

I proceed as follows: First, I briefly discuss my own foray into the world of participant observation, which occurred during the 2012 Senate race in Montana. Second, I consider two important methodological considerations raised by Fenno: access and participation as an observer. Third, I relate these two issues to a final consideration: the development of social media and the consequences of this for the participant observation enterprise. Finally, I show the perils of social science divorced from context, as demonstrated by the recent Stanford-Dartmouth mailer scandal. I conclude with not just a plea for us to pick up where Fenno has left off, but by suggesting that more thinking like a participant observer would benefit the discipline as whole by reminding us of our ethical obligations as researchers to each other, and to the political community that we study…(More)”.

Understanding Data Use: Building M&E Systems that Empower Users

Curated on Posted on by Stefaan Verhulst

Paper by Susan Stout, Vinisha Bhatia, and Paige Kirby: “We know that Monitoring and Evaluation (M&E) aims to support accountability and learning, in order to drive better outcomes…The paper, Understanding Data Use: Building M&E Systems that Empower Users, emphasizes how critical it is for decision makers to consider users’ decision space – from the institutional all the way to technical levels – in achieving data uptake.

Specifically, we call on smart mapping of this decision space – what do intended M&E users need, and what institutional factors shape those needs? With this understanding, we can better anticipate what types of data are most useful, and invest in systems to support data-driven decision making and better outcomes.

Mapping decision space is essential to understanding M&E data use. And as we’ve explored before, the development community has the opportunity to unlock existing resources to access more and better data that fits the needs of development actors to meet the SDGs….(More)”.

The rise of policy innovation labs: A catalog of policy innovation labs across Canada

Curated on Posted on by Stefaan Verhulst

Report by the Centre for Policy Innovation and Public Engagement (CPIPE): “In recent years, governments all over the world have been embracing new and innovative ways to develop public policies and design public services, from crowdsourcing to human-centred design thinking. This trend in government innovation has led to the rise of the Policy Innovation Lab (PIL): individual units, both inside and outside of government, that apply the traditional principles of scientific laboratories – experimentation, testing, and measurement – to social problems.

PILs are an increasingly important development in public policy making, with a variety of methods and approaches to building relationships between governments, organizations, and citizens, and generating ideas and designing policy. Yet, these labs are under-researched: many are established without a full understanding of their role and value to the policy community. We aim to address this knowledge gap, and create opportunities where policy innovators can make connections with their peers and learn about the current practices and applications of policy innovation from one another.

This report identifies the innovation labs in Canada, profiling their methodologies, projects, and partners, mapping the policy innovation landscape across the country. Each one-page summary provides a profile for each lab, and highlights the existing innovation practices and networks in the public, academic, non-profit, and private sectors, and identifies methodological and ideological trends across the different labs and networks.

This report is the first of its kind in North America. In this highly dynamic space, new labs are emerging and disappearing all the time. The purpose of this report is to put a spotlight on policy innovations and their successes, and to build and strengthen connections between researchers, policymakers, and policy innovators. Through a strengthened and sustained community of practice, we hope to see governments continue to embrace new approaches for effective policymaking…(More)”.

Better ways to measure the new economy

Curated on Posted on by Stefaan Verhulst

Valerie Hellinghausen and Evan Absher at Kauffman Foundation: “The old measure of “jobs numbers” as an economic indicator is shifting to new metrics to measure a new economy.

With more communities embracing inclusive entrepreneurial ecosystems as the new model of economic development, entrepreneurs, ecosystem builders, and government agencies – at all levels – need to work together on data-driven initiatives. While established measures still have a place, new metrics have the potential to deliver the timely and granular information that is more useful at the local level….

Three better ways to measure the new economy:

  1. National and local datasets:Numbers used to discuss the economy are national level and usually not very timely. These numbers are useful to understand large trends, but fail to capture local realities. One way to better measure local economies is to use local administrative datasets. There are many obstacles with this approach, but the idea is gaining interest. Data infrastructure, policies, and projects are building connections between local and national agencies. Joining different levels of government data will provide national scale and local specificity.
  1. Private and public data:The words private and public typically reflect privacy issues, but there is another public and private dimension. Public institutions possess vast amounts of data, but so do private companies. For instance, sites like PayPal, Square, Amazon, and Etsy possess data that could provide real-time assessment of an individual company’s financial health. The concept of credit and risk could be expanded to benefit those currently underserved, if combined with local administrative information like tax, wage, and banking data. Fair and open use of private data could open credit to currently underfunded entrepreneurs.
  1. New metrics:Developing connections between different datasets will result in new metrics of entrepreneurial activity: metrics that measure human connection, social capital, community creativity, and quality of life. Metrics that capture economic activity at the community level and in real time. For example, the Kauffman Foundation has funded research that uses labor data from private job-listing sites to better understand the match between the workforce entrepreneurs need and the workforce available within the immediate community. But new metrics are not enough, they must connect to the final goal of economic independence. Using new metrics to help ecosystems understand how policies and programs impact entrepreneurship is the final step to measuring local economies….(More)”.

An Overview of National AI Strategies

Curated on Posted on by Stefaan Verhulst

Medium Article by Tim Dutton: “The race to become the global leader in artificial intelligence (AI) has officially begun. In the past fifteen months, Canada, China, Denmark, the EU Commission, Finland, France, India, Italy, Japan, Mexico, the Nordic-Baltic region, Singapore, South Korea, Sweden, Taiwan, the UAE, and the UK have all released strategies to promote the use and development of AI. No two strategies are alike, with each focusing on different aspects of AI policy: scientific research, talent development, skills and education, public and private sector adoption, ethics and inclusion, standards and regulations, and data and digital infrastructure.

This article summarizes the key policies and goals of each strategy, as well as related policies and initiatives that have announced since the release of the initial strategies. It also includes countries that have announced their intention to develop a strategy or have related AI policies in place….(More)”.

‘To own or not to own?’ A study on the determinants and consequences of alternative intellectual property rights arrangements in crowdsourcing for innovation contests

Curated on Posted on by Stefaan Verhulst

Paper by Nuran Acur, Mariangela Piazza and Giovanni Perrone: “Firms are increasingly engaging in crowdsourcing for innovation to access new knowledge beyond their boundaries; however, scholars are no closer to understanding what guides seeker firms in deciding the level at which to acquire rights from solvers and the effect that this decision has on the performance of crowdsourcing contests.

Integrating Property Rights Theory and the problem solving perspective whist leveraging exploratory interviews and observations, we build a theoretical framework to examine how specific attributes of the technical problem broadcast affect the seekers’ choice between alternative intellectual property rights (IPR) arrangements that call for acquiring or licensing‐in IPR from external solvers (i.e. with high and low degrees of ownership respectively). Each technical problem differs in the knowledge required to be solved as well as in the stage of development it occurs of the innovation process and seeker firms pay great attention to such characteristics when deciding about the IPR arrangement they choose for their contests.

In addition, we analyze how this choice between acquiring and licensing‐in IPR, in turn, influences the performance of the contest. We empirically test our hypotheses analyzing a unique dataset of 729 challenges broadcast on the InnoCentive platform from 2010 to 2016. Our results indicate that challenges related to technical problems in later stages of the innovation process are positively related to the seekers’ preference toward IPR arrangements with a high level of ownership, while technical problems involving a higher number of knowledge domains are not.

Moreover, we found that IPR arrangements with a high level of ownership negatively affect solvers’ participation and that IPR arrangement plays a mediating role between the attributes of the technical problem and the solvers’ self‐selection process. Our article contributes to the open innovation and crowdsourcing literature and provides practical implications for both managers and contest organizers….(More)”.

Trust, Security, and Privacy in Crowdsourcing

Curated on Posted on by Stefaan Verhulst

Guest Editorial to Special Issue of IEEE Internet of Things Journal: “As we become increasingly reliant on intelligent, interconnected devices in every aspect of our lives, critical trust, security, and privacy concerns are raised as well.

First, the sensing data provided by individual participants is not always reliable. It may be noisy or even faked due to various reasons, such as poor sensor quality, lack of sensor calibration, background noise, context impact, mobility, incomplete view of observations, or malicious attacks. The crowdsourcing applications should be able to evaluate the trustworthiness of collected data in order to filter out the noisy and fake data that may disturb or intrude a crowdsourcing system. Second, providing data (e.g., photographs taken with personal mobile devices) or using IoT applications may compromise data providers’ personal data privacy (e.g., location, trajectory, and activity privacy) and identity privacy. Therefore, it becomes essential to assess the trust of the data while preserving the data providers’ privacy. Third, data analytics and mining in crowdsourcing may disclose the privacy of data providers or related entities to unauthorized parities, which lowers the willingness of participants to contribute to the crowdsourcing system, impacts system acceptance, and greatly impedes its further development. Fourth, the identities of data providers could be forged by malicious attackers to intrude the whole crowdsourcing system. In this context, trust, security, and privacy start to attract a special attention in order to achieve high quality of service in each step of crowdsourcing with regard to data collection, transmission, selection, processing, analysis and mining, as well as utilization.

Trust, security, and privacy in crowdsourcing receives increasing attention. Many methods have been proposed to protect privacy in the process of data collection and processing. For example, data perturbation can be adopted to hide the real data values during data collection. When preprocessing the collected data, data anonymization (e.g., k-anonymization) and fusion can be applied to break the links between the data and their sources/providers. In application layer, anonymity is used to mask the real identities of data sources/providers. To enable privacy-preserving data mining, secure multiparty computation (SMC) and homomorphic encryption provide options for protecting raw data when multiple parties jointly run a data mining algorithm. Through cryptographic techniques, no party knows anything else than its own input and expected results. For data truth discovery, applicable solutions include correlation-based data quality analysis and trust evaluation of data sources. But current solutions are still imperfect, incomprehensive, and inefficient….(More)”.

Data Science Thinking: The Next Scientific, Technological and Economic Revolution

Curated on Posted on by Stefaan Verhulst

Book by Longbing Cao: “This book explores answers to the fundamental questions driving the research, innovation and practices of the latest revolution in scientific, technological and economic development: how does data science transform existing science, technology, industry, economy, profession and education?  How does one remain competitive in the data science field? What is responsible for shaping the mindset and skillset of data scientists?

Data Science Thinking paints a comprehensive picture of data science as a new scientific paradigm from the scientific evolution perspective, as data science thinking from the scientific-thinking perspective, as a trans-disciplinary science from the disciplinary perspective, and as a new profession and economy from the business perspective.

The topics cover an extremely wide spectrum of essential and relevant aspects of data science, spanning its evolution, concepts, thinking, challenges, discipline, and foundation, all the way to industrialization, profession, education, and the vast array of opportunities that data science offers. The book’s three parts each detail layers of these different aspects….(More)”.

The Risks of Dangerous Dashboards in Basic Education

Curated on Posted on by Stefaan Verhulst

Lant Pritchett at the Center for Global Development: “On June 1, 2009 Air France flight 447 from Rio de Janeiro to Paris crashed into the Atlantic Ocean killing all 228 people on board. While the Airbus 330 was flying on auto-pilot, the different speed indicators received by the on-board navigation computers started to give conflicting speeds, almost certainly because the pitot tubes responsible for measuring air speed had iced over. Since the auto-pilot could not resolve conflicting signals and hence did not know how fast the plane was actually going, it turned control of the plane over to the two first officers (the captain was out of the cockpit). Subsequent flight simulator trials replicating the conditions of the flight conclude that had the pilots done nothing at all everyone would have lived—nothing was actually wrong; only the indicators were faulty, not the actual speed. But, tragically, the pilots didn’t do nothing….

What is the connection to education?

Many countries’ systems of basic education are in “stall” condition.

A recent paper of Beatty et al. (2018) uses information from the Indonesia Family Life Survey, a representative household survey that has been carried out in several waves with the same individuals since 2000 and contains information on whether individuals can answer simple arithmetic questions. Figure 1, showing the relationship between the level of schooling and the probability of answering a typical question correctly, has two shocking results.

First, the difference in the likelihood a person can answer a simple mathematics question correctly differs by only 20 percent between individuals who have completed less than primary school (<PS)—who can answer correctly (adjusted for guessing) about 20 percent of the time—and those who have completed senior secondary school or more (>=SSS), who answer correctly only about 40 percent of the time. These are simple multiple choice questions like whether 56/84 is the same fraction as (can be reduced to) 2/3, and whether 1/3-1/6 equals 1/6. This means that in an entire year of schooling, less than 2 additional children per 100 gain the ability to answer simple arithmetic questions.

Second, this incredibly poor performance in 2000 got worse by 2014. …

What has this got to do with education dashboards? The way large bureaucracies prefer to work is to specify process compliance and inputs and then measure those as a means of driving performance. This logistical mode of managing an organization works best when both process compliance and inputs are easily “observable” in the economist’s sense of easily verifiable, contractible, adjudicated. This leads to attention to processes and inputs that are “thin” in the Clifford Geertz sense (adopted by James Scott as his primary definition of how a “high modern” bureaucracy and hence the state “sees” the world). So in education one would specify easily-observable inputs like textbook availability, class size, school infrastructure. Even if one were talking about “quality” of schooling, a large bureaucracy would want this too reduced to “thin” indicators, like the fraction of teachers with a given type of formal degree, or process compliance measures, like whether teachers were hired based on some formal assessment.

Those involved in schooling can then become obsessed with their dashboards and the “thin” progress that is being tracked and easily ignore the loud warning signals saying: Stall!…(More)”.