The Bad Pupil

Curated on Posted on by Stefaan Verhulst

CCCBLab: “In recent years we have been witnessing a constant trickle of news on artificial intelligence, machine learning and computer vision. We are told that machines learn, see, create… and all this builds up a discourse based on novelty, on a possible future and on a series of worries and hopes. It is difficult, sometimes, to figure out in this landscape which are real developments, and which are fantasies or warnings. And, undoubtedly, this fog that surrounds it forms part of the power that we grant, both in the present and on credit, to these tools, and of the negative and positive concerns that they arouse in us. Many of these discourses may fall into the field of false debates or, at least, of the return of old debates. Thus, in the classical artistic field, associated with the discourse on creation and authorship, there is discussion regarding the entity to be awarded to the images created with these tools. (Yet wasn’t the argument against photography in art that it was an image created automatically and without human participation? And wasn’t that also an argument in favour of taking it and using it to put an end to a certain idea of art?)

Metaphors are essential in the discourse on all digital tools and the power that they have. Are expressions such as “intelligence”, “vision”, “learning”, “neural” and the entire range of similar words the most adequate for defining these types of tools? Probably not, above all if their metaphorical nature is sidestepped. We would not understand them in the same way if we called them tools of probabilistic classification or if instead of saying that an artificial intelligence “has painted” a Rembrandt, we said that it has produced a statistical reproduction of his style (something which is still surprising, and to be celebrated, of course). These names construct an entity for these tools that endows them with a supposed autonomy and independence upon which their future authority is based.

Because that is what it’s about in many discourses: constructing a characterisation that legitimises an objective or non-human capacity in data analysis….

We now find ourselves in what is, probably, the point of the first cultural reception of these tools. Of their development in fields of research and applications that have already been derived, we are moving on to their presence in the public discourse. It is in this situation and context, where we do not fully know the breadth and characteristics of these technologies (meaning fears are more abstract and diffuse and, thus, more present and powerful), when it is especially important to understand what we are talking about, to appropriate the tools and to intervene in the discourses. Before their possibilities are restricted and solidified until they seem indisputable, it is necessary to experiment with them and reflect on them; taking advantage of the fact that we can still easily perceive them as in creation, malleable and open.

In our projects The Bad Pupil. Critical pedagogy for artificial intelligences and Latent Spaces. Machinic Imaginations we have tried to approach to these tools and their imaginary. In the statement of intentions of the former, we expressed our desire, in the face of the regulatory context and the metaphor of machine learning, to defend the bad pupil as one who escapes the norm. And also how, faced with an artificial intelligence that seeks to replicate the human on inhuman scales, it is necessary to defend and construct a non-mimetic one that produces unexpected relations and images.

Fragment of De zeven werken van barmhartigheid, Meester van Alkmaar, 1504 (Rijksmuseum, Amsterdam) analysed with YOLO9000 | The Bad Pupil - Estampa

Fragment of De zeven werken van barmhartigheid, Meester van Alkmaar, 1504 (Rijksmuseum, Amsterdam) analysed with YOLO9000 | The Bad Pupil – Estampa

Both projects are also attempts to appropriate these tools, which means, first of all, escaping industrial barriers and their standards. In this field in which mass data are an asset within reach of big companies, employing quantitively poor datasets and non-industrial calculation potentials is not just a need but a demand….(More)”.

Public Interest Technology University Network

Curated on Posted on by Stefaan Verhulst

About: “The Public Interest Technology Universities Network is a partnership that fosters collaboration between 21 universities and colleges committed to building the nascent field of public interest technology and growing a new generation of civic-minded technologists. Through the development of curricula, research agendas, and experiential learning programs in the public interest technology space, these universities are trying innovative tactics to produce graduates with multiple fluencies at the intersection of technology and policy. By joining PIT-UN, members commit to field building on campus. Members may choose to focus on some or all of these elements, in addition to other initiatives they deem relevant to establishing public interest technology on campus:

  1. Support curriculum and faculty development to enable interdisciplinary and cross-disciplinary education of students, so they can critically assess the ethical, political, and societal implications of new technologies, and design technologies in service of the public good.
  2. Develop experiential learning opportunities such as clinics, fellowships, apprenticeships, and internship, with public and private sector partners in the public interest technology space.
  3. Find ways to support graduates who pursue careers working in public interest technology, recognizing that financial considerations may make careers in this area unaffordable to many.
  4. Create mechanisms for faculty to receive recognition for the research, curriculum development, teaching, and service work needed to build public interest technology as an arena of inquiry.
  5. Provide institutional data that will allow us to measure the effectiveness of our interventions in helping to develop the field of public interest technology….(More)”.

Applying behavioral insights to improve postsecondary education outcomes

Curated on Posted on by Stefaan Verhulst

Brookings: “Policymakers under President Obama implemented behaviorally-informed policies to improve college access, completion, and affordability. Given the complexity of the college application process, many of these policies aimed to simplify college and financial aid application processes and reduce informational barriers that students face when evaluating college options. Katharine Meyer and Kelly Ochs Rosinger summarize empirical evidence on these policies and conclude that behaviorally-informed policies play an important role, especially as supplements to (rather than replacements for) broader structural changes. For example, recent changes in the FAFSA filing timeline provided students with more time to complete the form. But this large shift may be more effective in changing behavior when accompanied by informational campaigns and nudges that improve students’ understanding of the new system. Governments and colleges can leverage behavioral science to increase awareness of student support services if more structural policy changes occur to provide the services in the first place….(More)”.

Conceptualising the Digital University

Curated on Posted on by Stefaan Verhulst

Book edited by Bill Johnston, Sheila MacNeill and Keith Smyth: “Despite the increasing ubiquity of the term, the concept of the digital university remains diffuse and indeterminate. This book examines what the term ‘digital university’ should encapsulate and the resulting challenges, possibilities and implications that digital technology and practice brings to higher education. Critiquing the current state of definition of the digital university construct, the authors propose a more holistic, integrated account that acknowledges the inherent diffuseness of the concept. The authors also question the extent to which digital technologies and practices can allow us to re-think the location of universities and curricula; and how they can extend higher education as a public good within the current wider political context. Framed inside a critical pedagogy perspective, this volume debates the role of the university in fostering the learning environments, skills and capabilities needed for critical engagement, active open participation and reflection in the digital age. This pioneering volume will be of interest and value to students and scholars of digital education, as well as policy makers and practitioners….(More)”

Index: Open Data

Curated on Posted on by Andrew Young

By Alexandra Shaw, Michelle Winowatan, Andrew Young, and Stefaan Verhulst

The Living Library Index – inspired by the Harper’s Index – provides important statistics and highlights global trends in governance innovation. This installment focuses on open data and was originally published in 2018.

Value and Impact

  • The projected year at which all 28+ EU member countries will have a fully operating open data portal: 2020

  • Between 2016 and 2020, the market size of open data in Europe is expected to increase by 36.9%, and reach this value by 2020: EUR 75.7 billion

Public Views on and Use of Open Government Data

  • Number of Americans who do not trust the federal government or social media sites to protect their data: Approximately 50%

  • Key findings from The Economist Intelligence Unit report on Open Government Data Demand:

    • Percentage of respondents who say the key reason why governments open up their data is to create greater trust between the government and citizens: 70%

    • Percentage of respondents who say OGD plays an important role in improving lives of citizens: 78%

    • Percentage of respondents who say OGD helps with daily decision making especially for transportation, education, environment: 53%

    • Percentage of respondents who cite lack of awareness about OGD and its potential use and benefits as the greatest barrier to usage: 50%

    • Percentage of respondents who say they lack access to usable and relevant data: 31%

    • Percentage of respondents who think they don’t have sufficient technical skills to use open government data: 25%

    • Percentage of respondents who feel the number of OGD apps available is insufficient, indicating an opportunity for app developers: 20%

    • Percentage of respondents who say OGD has the potential to generate economic value and new business opportunity: 61%

    • Percentage of respondents who say they don’t trust governments to keep data safe, protected, and anonymized: 19%

Efforts and Involvement

  • Time that’s passed since open government advocates convened to create a set of principles for open government data – the instance that started the open data government movement: 10 years

  • Countries participating in the Open Government Partnership today: 79 OGP participating countries and 20 subnational governments

  • Percentage of “open data readiness” in Europe according to European Data Portal: 72%

    • Open data readiness consists of four indicators which are presence of policy, national coordination, licensing norms, and use of data.

  • Number of U.S. cities with Open Data portals: 27

  • Number of governments who have adopted the International Open Data Charter: 62

  • Number of non-state organizations endorsing the International Open Data Charter: 57

  • Number of countries analyzed by the Open Data Index: 94

  • Number of Latin American countries that do not have open data portals as of 2017: 4 total – Belize, Guatemala, Honduras and Nicaragua

  • Number of cities participating in the Open Data Census: 39

Demand for Open Data

  • Open data demand measured by frequency of open government data use according to The Economist Intelligence Unit report:

    • Australia

      • Monthly: 15% of respondents

      • Quarterly: 22% of respondents

      • Annually: 10% of respondents

    • Finland

      • Monthly: 28% of respondents

      • Quarterly: 18% of respondents

      • Annually: 20% of respondents

    •  France

      • Monthly: 27% of respondents

      • Quarterly: 17% of respondents

      • Annually: 19% of respondents

        •  

    • India

      • Monthly: 29% of respondents

      • Quarterly: 20% of respondents

      • Annually: 10% of respondents

    • Singapore

      • Monthly: 28% of respondents

      • Quarterly: 15% of respondents

      • Annually: 17% of respondents 

    • UK

      • Monthly: 23% of respondents

      • Quarterly: 21% of respondents

      • Annually: 15% of respondents

    • US

      • Monthly: 16% of respondents

      • Quarterly: 15% of respondents

      • Annually: 20% of respondents

  • Number of FOIA requests received in the US for fiscal year 2017: 818,271

  • Number of FOIA request processed in the US for fiscal year 2017: 823,222

  • Distribution of FOIA requests in 2017 among top 5 agencies with highest number of request:

    • DHS: 45%

    • DOJ: 10%

    • NARA: 7%

    • DOD: 7%

    • HHS: 4%

Examining Datasets

  • Country with highest index score according to ODB Leaders Edition: Canada (76 out of 100)

  • Country with lowest index score according to ODB Leaders Edition: Sierra Leone (22 out of 100)

  • Number of datasets open in the top 30 governments according to ODB Leaders Edition: Fewer than 1 in 5

  • Average percentage of datasets that are open in the top 30 open data governments according to ODB Leaders Edition: 19%

  • Average percentage of datasets that are open in the top 30 open data governments according to ODB Leaders Edition by sector/subject:

    • Budget: 30%

    • Companies: 13%

    • Contracts: 27%

    • Crime: 17%

    • Education: 13%

    • Elections: 17%

    • Environment: 20%

    • Health: 17%

    • Land: 7%

    • Legislation: 13%

    • Maps: 20%

    • Spending: 13%

    • Statistics: 27%

    • Trade: 23%

    • Transport: 30%

  • Percentage of countries that release data on government spending according to ODB Leaders Edition: 13%

  • Percentage of government data that is updated at regular intervals according to ODB Leaders Edition: 74%

  • Number of datasets available through:

  • Number of datasets classed as “open” in 94 places worldwide analyzed by the Open Data Index: 11%

  • Percentage of open datasets in the Caribbean, according to Open Data Census: 7%

  • Number of companies whose data is available through OpenCorporates: 158,589,950

City Open Data

  • New York City

  • Singapore

    • Number of datasets published in Singapore: 1,480

    • Percentage of datasets with standardized format: 35%

    • Percentage of datasets made as raw as possible: 25%

  • Barcelona

    • Number of datasets published in Barcelona: 443

    • Open data demand in Barcelona measured by:

      • Number of unique sessions in the month of September 2018: 5,401

    • Quality of datasets published in Barcelona according to Tim Berners Lee 5-star Open Data: 3 stars

  • London

    • Number of datasets published in London: 762

    • Number of data requests since October 2014: 325

  • Bandung

    • Number of datasets published in Bandung: 1,417

  • Buenos Aires

    • Number of datasets published in Buenos Aires: 216

  • Dubai

    • Number of datasets published in Dubai: 267

  • Melbourne

    • Number of datasets published in Melbourne: 199

Sources

  • About OGP, Open Government Partnership. 2018.  

Seven design principles for using blockchain for social impact

Curated on Posted on by Stefaan Verhulst

Stefaan Verhulst at Apolitical: “2018 will probably be remembered as the bust of the blockchain hype. Yet even as crypto currencies continue to sink in value and popular interest, the potential of using blockchain technologies to achieve social ends remains important to consider but poorly understood.

In 2019, business will continue to explore blockchain for sectors as disparate as finance, agriculture, logistics and healthcare. Policymakers and social innovators should also leverage 2019 to become more sophisticated about blockchain’s real promise, limitations  and current practice.

In a recent report I prepared with Andrew Young, with the support of the Rockefeller Foundation, we looked at the potential risks and challenges of using blockchain for social change — or “Blockchan.ge.” A number of implementations and platforms are already demonstrating potential social impact.

The technology is now being used to address issues as varied as homelessness in New York City, the Rohingya crisis in Myanmar and government corruption around the world.

In an illustration of the breadth of current experimentation, Stanford’s Center for Social Innovation recently analysed and mapped nearly 200 organisations and projects trying to create positive social change using blockchain. Likewise, the GovLab is developing a mapping of blockchange implementations across regions and topic areas; it currently contains 60 entries.

All these examples provide impressive — and hopeful — proof of concept. Yet despite the very clear potential of blockchain, there has been little systematic analysis. For what types of social impact is it best suited? Under what conditions is it most likely to lead to real social change? What challenges does blockchain face, what risks does it pose and how should these be confronted and mitigated?

These are just some of the questions our report, which builds its analysis on 10 case studies assembled through original research, seeks to address.

While the report is focused on identity management, it contains a number of lessons and insights that are applicable more generally to the subject of blockchange.

In particular, it contains seven design principles that can guide individuals or organisations considering the use of blockchain for social impact. We call these the Genesis principles, and they are outlined at the end of this article…(More)”.

Public Attitudes Toward Computer Algorithms

Curated on Posted on by Stefaan Verhulst

Aaron Smith at the Pew Research Center: “Algorithms are all around us, utilizing massive stores of data and complex analytics to make decisions with often significant impacts on humans. They recommend books and movies for us to read and watch, surface news stories they think we might find relevant, estimate the likelihood that a tumor is cancerous and predict whether someone might be a criminal or a worthwhile credit risk. But despite the growing presence of algorithms in many aspects of daily life, a Pew Research Center survey of U.S. adults finds that the public is frequently skeptical of these tools when used in various real-life situations.

This skepticism spans several dimensions. At a broad level, 58% of Americans feel that computer programs will always reflect some level of human bias – although 40% think these programs can be designed in a way that is bias-free. And in various contexts, the public worries that these tools might violate privacy, fail to capture the nuance of complex situations, or simply put the people they are evaluating in an unfair situation. Public perceptions of algorithmic decision-making are also often highly contextual. The survey shows that otherwise similar technologies can be viewed with support or suspicion depending on the circumstances or on the tasks they are assigned to do….

The following are among the major findings.

The public expresses broad concerns about the fairness and acceptability of using computers for decision-making in situations with important real-world consequences

Majorities of Americans find it unacceptable to use algorithms to make decisions with real-world consequences for humans

By and large, the public views these examples of algorithmic decision-making as unfair to the people the computer-based systems are evaluating. Most notably, only around one-third of Americans think that the video job interview and personal finance score algorithms would be fair to job applicants and consumers. When asked directly whether they think the use of these algorithms is acceptable, a majority of the public says that they are not acceptable. Two-thirds of Americans (68%) find the personal finance score algorithm unacceptable, and 67% say the computer-aided video job analysis algorithm is unacceptable….

Attitudes toward algorithmic decision-making can depend heavily on context

Despite the consistencies in some of these responses, the survey also highlights the ways in which Americans’ attitudes toward algorithmic decision-making can depend heavily on the context of those decisions and the characteristics of the people who might be affected….

When it comes to the algorithms that underpin the social media environment, users’ comfort level with sharing their personal information also depends heavily on how and why their data are being used. A 75% majority of social media users say they would be comfortable sharing their data with those sites if it were used to recommend events they might like to attend. But that share falls to just 37% if their data are being used to deliver messages from political campaigns.

Across age groups, social media users are comfortable with their data being used to recommend events - but wary of that data being used for political messaging

In other instances, different types of users offer divergent views about the collection and use of their personal data. For instance, about two-thirds of social media users younger than 50 find it acceptable for social media platforms to use their personal data to recommend connecting with people they might want to know. But that view is shared by fewer than half of users ages 65 and older….(More)”.

A Behavioral Economics Approach to Digitalisation

Curated on Posted on by Stefaan Verhulst

Paper by Dirk Beerbaum and Julia M. Puaschunder: “A growing body of academic research in the field of behavioural economics, political science and psychology demonstrate how an invisible hand can nudge people’s decisions towards a preferred option. Contrary to the assumptions of the neoclassical economics, supporters of nudging argue that people have problems coping with a complex world, because of their limited knowledge and their restricted rationality. Technological improvement in the age of information has increased the possibilities to control the innocent social media users or penalise private investors and reap the benefits of their existence in hidden persuasion and discrimination. Nudging enables nudgers to plunder the simple uneducated and uninformed citizen and investor, who is neither aware of the nudging strategies nor able to oversee the tactics used by the nudgers (Puaschunder 2017a, b; 2018a, b).

The nudgers are thereby legally protected by democratically assigned positions they hold. The law of motion of the nudging societies holds an unequal concentration of power of those who have access to compiled data and coding rules, relevant for political power and influencing the investor’s decision usefulness (Puaschunder 2017a, b; 2018a, b). This paper takes as a case the “transparency technology XBRL (eXtensible Business Reporting Language)” (Sunstein 2013, 20), which should make data more accessible as well as usable for private investors. It is part of the choice architecture on regulation by governments (Sunstein 2013). However, XBRL is bounded to a taxonomy (Piechocki and Felden 2007).

Considering theoretical literature and field research, a representation issue (Beerbaum, Piechocki and Weber 2017) for principles-based accounting taxonomies exists, which intelligent machines applying Artificial Intelligence (AI) (Mwilu, Prat and Comyn-Wattiau 2015) nudge to facilitate decision usefulness. This paper conceptualizes ethical questions arising from the taxonomy engineering based on machine learning systems: Should the objective of the coding rule be to support or to influence human decision making or rational artificiality? This paper therefore advocates for a democratisation of information, education and transparency about nudges and coding rules (Puaschunder 2017a, b; 2018a, b)…(More)”.

Better Ways to Communicate Hospital Data to Physicians

Curated on Posted on by Stefaan Verhulst

Scott Falk, John Cherf and Julie Schulz at the Harvard Business Review: “We recently conducted an in-depth study at Lumere to gain insight into physicians’ perceptions of clinical variation and the factors influencing their choices of drugs and devices. Based on a survey of 276 physicians, our study results show that it’s necessary to consistently and frequently share cost data and clinical evidence with physicians, regardless of whether they’re affiliated with or directly employed by a hospital….

There are multiple explanations as to why health system administrators have been slow to share data with physicians. The two most common challenges are difficulty obtaining accurate, clinically meaningful data and lack of knowledge among administrators about communicating data.

When it comes to obtaining accurate, meaningful data, the reality is that many health systems do not know where to start. Between disparate data-collection systems, varied physician needs, and an overwhelming array of available clinical evidence, it can be daunting to try to develop a robust, yet streamlined, approach.

As for the second problem, many administrators have simply not been trained to effectively communicate data. Health system leaders tend to be more comfortable talking about costs, but physicians generally focus on clinical outcomes. As a result, physicians frequently have follow-up questions that administrators interpret as pushback. It is important to understand what physicians need.

Determine the appropriate amount and type of data to share. Using evidence and data can foster respectful debate, provide honest education, and ultimately align teams.

Physicians are driven by their desire to improve patient outcomes and therefore want the total picture. This includes access to published evidence to help choose cost-effective drug and device alternatives without hurting outcomes. Health system administrators need to provide clinicians with access to a wide range of data (not only data about costs). Ensuring that physicians have a strong voice in determining which data to share will help create alignment and trust. A more nuanced value-based approach that accounts for important clinical and patient-centered outcomes (e.g., length of stay, post-operative recovery profile) combined with cost data may be the most effective solution.

While physicians generally report wanting more cost data, not all physicians have the experience and training to appropriately incorporate it into their decision making. Surveyed physicians who have had exposure to a range of cost data, data highlighting clinical variation, and practice guidelines generally found cost data more influential in their selection of drugs and devices, regardless of whether they shared in savings under value-based care models. This was particularly true for more veteran physicians and those with private-practice experience who have had greater exposure to managing cost information.

Health systems can play a key role in helping physicians use cost and quality data to make cost-effective decisions. We recommend that health systems identify a centralized data/analytics department that includes representatives of both quality-improvement teams and technology/informatics to own the process of streamlining, analyzing, and disseminating data.

Compare data based on contemporary evidence-based guidelines. Physicians would like to incorporate reliable data into their decision-making when selecting drugs and devices. In our survey, 54% of respondents reported that it was either “extremely important” or “very important” that hospitals use peer-reviewed literature and clinical evidence to support the selection of medical devices. Further, 56% of respondents said it was “extremely important” or “very important” that physicians be involved in using data to develop clinical protocols, guidelines, and best practices….(More)”.

Positive deviance, big data, and development: A systematic literature review

Curated on Posted on by Stefaan Verhulst

Paper by Basma Albanna and Richard Heeks: “Positive deviance is a growing approach in international development that identifies those within a population who are outperforming their peers in some way, eg, children in low‐income families who are well nourished when those around them are not. Analysing and then disseminating the behaviours and other factors underpinning positive deviance are demonstrably effective in delivering development results.

However, positive deviance faces a number of challenges that are restricting its diffusion. In this paper, using a systematic literature review, we analyse the current state of positive deviance and the potential for big data to address the challenges facing positive deviance. From this, we evaluate the promise of “big data‐based positive deviance”: This would analyse typical sources of big data in developing countries—mobile phone records, social media, remote sensing data, etc—to identify both positive deviants and the factors underpinning their superior performance.

While big data cannot solve all the challenges facing positive deviance as a development tool, they could reduce time, cost, and effort; identify positive deviants in new or better ways; and enable positive deviance to break out of its current preoccupation with public health into domains such as agriculture, education, and urban planning. In turn, positive deviance could provide a new and systematic basis for extracting real‐world development impacts from big data…(More)”.