Cary Coglianese: “The widespread public angst that surfaced in the 2016 presidential election revealed how many Americans believe their government has become badly broken. Given the serious problems that continue to persist in society — crime, illiteracy, unemployment, poverty, discrimination, to name a few — widespread beliefs in a governmental breakdown are understandable. Yet such a breakdown is actually far from self-evident. In this paper, I explain how diagnoses of governmental performance depend on the perspective from which current conditions in the country are viewed. Certainly when judged against a standard of perfection, America has a long way to go. But perfection is no meaningful basis upon which to conclude that government has broken down. I offer and assess three alternative, more realistic benchmarks of government’s performance: (1) reliance on a standard of acceptable imperfection; (2) comparisons with other countries or time periods; and (3) the use of counterfactual inferences. Viewed from these perspectives, the notion of an irreparable governmental failure in the United States becomes quite questionable. Although serious economic and social shortcomings certainly exist, the nation’s strong economy and steadily improving living conditions in recent decades simply could not have occurred if government were not functioning well. Rather than embracing despair and giving in to cynicism and resignation, citizens and their leaders would do better to treat the nation’s problems as conditions of disrepair needing continued democratic engagement. It remains possible to achieve greater justice and better economic and social conditions for all — but only if we, the people, do not give up on the pursuit of these goals….(More)”
What’s wrong with big data?
James Bridle in the New Humanist: “In a 2008 article in Wired magazine entitled “The End of Theory”, Chris Anderson argued that the vast amounts of data now available to researchers made the traditional scientific process obsolete. No longer would they need to build models of the world and test them against sampled data. Instead, the complexities of huge and totalising datasets would be processed by immense computing clusters to produce truth itself: “With enough data, the numbers speak for themselves.” As an example, Anderson cited Google’s translation algorithms which, with no knowledge of the underlying structures of languages, were capable of inferring the relationship between them using extensive corpora of translated texts. He extended this approach to genomics, neurology and physics, where scientists are increasingly turning to massive computation to make sense of the volumes of information they have gathered about complex systems. In the age of big data, he argued, “Correlation is enough. We can stop looking for models.”
This belief in the power of data, of technology untrammelled by petty human worldviews, is the practical cousin of more metaphysical assertions. A belief in the unquestionability of data leads directly to a belief in the truth of data-derived assertions. And if data contains truth, then it will, without moral intervention, produce better outcomes. Speaking at Google’s private London Zeitgeist conference in 2013, Eric Schmidt, Google Chairman, asserted that “if they had had cellphones in Rwanda in 1994, the genocide would not have happened.” Schmidt’s claim was that technological visibility – the rendering of events and actions legible to everyone – would change the character of those actions. Not only is this statement historically inaccurate (there was plenty of evidence available of what was occurring during the genocide from UN officials, US satellite photographs and other sources), it’s also demonstrably untrue. Analysis of unrest in Kenya in 2007, when over 1,000 people were killed in ethnic conflicts, showed that mobile phones not only spread but accelerated the violence. But you don’t need to look to such extreme examples to see how a belief in technological determinism underlies much of our thinking and reasoning about the world.
“Big data” is not merely a business buzzword, but a way of seeing the world. Driven by technology, markets and politics, it has come to determine much of our thinking, but it is flawed and dangerous. It runs counter to our actual findings when we employ such technologies honestly and with the full understanding of their workings and capabilities. This over-reliance on data, which I call “quantified thinking”, has come to undermine our ability to reason meaningfully about the world, and its effects can be seen across multiple domains.
The assertion is hardly new. Writing in the Dialectic of Enlightenment in 1947, Theodor Adorno and Max Horkheimer decried “the present triumph of the factual mentality” – the predecessor to quantified thinking – and succinctly analysed the big data fallacy, set out by Anderson above. “It does not work by images or concepts, by the fortunate insights, but refers to method, the exploitation of others’ work, and capital … What men want to learn from nature is how to use it in order wholly to dominate it and other men. That is the only aim.” What is different in our own time is that we have built a world-spanning network of communication and computation to test this assertion. While it occasionally engenders entirely new forms of behaviour and interaction, the network most often shows to us with startling clarity the relationships and tendencies which have been latent or occluded until now. In the face of the increased standardisation of knowledge, it becomes harder and harder to argue against quantified thinking, because the advances of technology have been conjoined with the scientific method and social progress. But as I hope to show, technology ultimately reveals its limitations….
“Eroom’s law” – Moore’s law backwards – was recently formulated to describe a problem in pharmacology. Drug discovery has been getting more expensive. Since the 1950s the number of drugs approved for use in human patients per billion US dollars spent on research and development has halved every nine years. This problem has long perplexed researchers. According to the principles of technological growth, the trend should be in the opposite direction. In a 2012 paper in Nature entitled “Diagnosing the decline in pharmaceutical R&D efficiency” the authors propose and investigate several possible causes for this. They begin with social and physical influences, such as increased regulation, increased expectations and the exhaustion of easy targets (the “low hanging fruit” problem). Each of these are – with qualifications – disposed of, leaving open the question of the discovery process itself….(More)
Open Data Collection (PLOS)
Daniella Lowenberg, Amy Ross, Emma Ganley at PLOS: “In the spirit of Open Con and highlighting the state of Open Data, PLOS is proud to release our Open Data Collection. The many values of Open Data are becoming increasingly apparent, and as a result, we are seeing an adoption of Open Data policies across publishers, funders and organizations. Open Data has proven a fantastic tool to help evaluate the replicability of published research, and even politicians are taking a stand in favor of Open Data as a mechanism to advance science rapidly. In March of 2014, PLOS updated our Data Policy to reflect the need for the underlying data to be as open as the paper itself resulting in complete transparency of the research. Two and-a-half years later, we have seen over 60,000 published papers with open data sets and an increase in submissions reflecting open data practices and policies….
To create this Open Data Collection, we exhaustively searched for relevant articles published across PLOS that discuss open data in some way. Then, in collaboration with our external advisor, Melissa Haendel, we have selected 26 of those articles with the aim to highlight a broad scope of research articles, guidelines, and commentaries about data sharing, data practices, and data policies from different research fields. Melissa has written an engaging blog post detailing the rubric and reasons behind her selections….(More)”
Understanding the four types of AI, from reactive robots to self-aware beings
The Conversation: “…We need to overcome the boundaries that define the four different types of artificial intelligence, the barriers that separate machines from us – and us from them.
Type I AI: Reactive machines
The most basic types of AI systems are purely reactive, and have the ability neither to form memories nor to use past experiences to inform current decisions. Deep Blue, IBM’s chess-playing supercomputer, which beat international grandmaster Garry Kasparov in the late 1990s, is the perfect example of this type of machine.
Deep Blue can identify the pieces on a chess board and know how each moves. It can make predictions about what moves might be next for it and its opponent. And it can choose the most optimal moves from among the possibilities.
But it doesn’t have any concept of the past, nor any memory of what has happened before. Apart from a rarely used chess-specific rule against repeating the same move three times, Deep Blue ignores everything before the present moment. All it does is look at the pieces on the chess board as it stands right now, and choose from possible next moves.
This type of intelligence involves the computer perceiving the world directly and acting on what it sees. It doesn’t rely on an internal concept of the world. In a seminal paper, AI researcher Rodney Brooks argued that we should only build machines like this. His main reason was that people are not very good at programming accurate simulated worlds for computers to use, what is called in AI scholarship a “representation” of the world….
Type II AI: Limited memory
This Type II class contains machines can look into the past. Self-driving cars do some of this already. For example, they observe other cars’ speed and direction. That can’t be done in a just one moment, but rather requires identifying specific objects and monitoring them over time.
These observations are added to the self-driving cars’ preprogrammed representations of the world, which also include lane markings, traffic lights and other important elements, like curves in the road. They’re included when the car decides when to change lanes, to avoid cutting off another driver or being hit by a nearby car.
But these simple pieces of information about the past are only transient. They aren’t saved as part of the car’s library of experience it can learn from, the way human drivers compile experience over years behind the wheel…;
Type III AI: Theory of mind
We might stop here, and call this point the important divide between the machines we have and the machines we will build in the future. However, it is better to be more specific to discuss the types of representations machines need to form, and what they need to be about.
Machines in the next, more advanced, class not only form representations about the world, but also about other agents or entities in the world. In psychology, this is called “theory of mind” – the understanding that people, creatures and objects in the world can have thoughts and emotions that affect their own behavior.
This is crucial to how we humans formed societies, because they allowed us to have social interactions. Without understanding each other’s motives and intentions, and without taking into account what somebody else knows either about me or the environment, working together is at best difficult, at worst impossible.
If AI systems are indeed ever to walk among us, they’ll have to be able to understand that each of us has thoughts and feelings and expectations for how we’ll be treated. And they’ll have to adjust their behavior accordingly.
Type IV AI: Self-awareness
The final step of AI development is to build systems that can form representations about themselves. Ultimately, we AI researchers will have to not only understand consciousness, but build machines that have it….
While we are probably far from creating machines that are self-aware, we should focus our efforts toward understanding memory, learning and the ability to base decisions on past experiences….(More)”
The Cost of Cooperating
If you think about the puzzle of cooperation being “why should I incur a personal cost of time or money or effort in order to do something that’s going to benefit other people and not me?” the general answer is that if you can create future consequences for present behavior, that can create an incentive to cooperate. Cooperation requires me to incur some costs now, but if I’m cooperating with someone who I’ll interact with again, it’s worth it for me to pay the cost of cooperating now in order to get the benefit of them cooperating with me in the future, as long as there’s a large enough likelihood that we’ll interact again.
Even if it’s with someone that I’m not going to interact with again, if other people are observing that interaction, then it affects my reputation. It can be worth paying the cost of cooperating in order to earn a good reputation, and to attract new interaction partners.
There’s a lot of evidence to show that this works. There are game theory models and computer simulations showing that if you build these kinds of future consequences, you can get either evolution to lead to cooperative agents dominating populations, and also learning and strategic reasoning leading to people cooperating. There are also lots of behavioral experiments supporting this. These are lab experiments where you bring people into the lab, give them money, and you have them engage in economic cooperation games where they choose whether to keep the money for themselves or to contribute it to a group project that benefits other people. If you make it so that future consequences exist in any of these various ways, it makes people more inclined to cooperate. Typically, it leads to cooperation paying off, and being the best-performing strategy.
In these situations, it’s not altruistic to be cooperative because the interactions are designed in a way that makes cooperating pay off. For example, we have a paper that shows that in the context of repeated interactions, there’s not any relationship between how altruistic people are and how much they cooperate. Basically, everybody cooperates, even the selfish people. Under certain situations, selfish people can even wind up cooperating more because they’re better at identifying that that’s what is going to pay off.
This general class of solutions to the cooperation problem boils down to creating future consequences, and therefore creating a self-interested motivation in the long run to be cooperative. Strategic cooperation is extremely important; it explains a lot of real-world cooperation. From an institution design perspective, it’s important for people to be thinking about how you set up the rules of interaction—interaction structures and incentive structures—in a way that makes working for the greater good a good strategy.
At the same time that this strategic cooperation is important, it’s also clearly the case that people often cooperate even when there’s not a self-interested motive to do so. That willingness to help strangers (or to not exploit them) is a core piece of well-functioning societies. It makes society much more efficient when you don’t constantly have to be watching your back, afraid that people are going to take advantage of you. If you can generally trust that other people are going to do the right thing and you’re going to do the right thing, it makes life much more socially efficient.
Strategic incentives can motivate people to cooperate, but people also keep cooperating even when there are not incentives to do so, at least to some extent. What motivates people to do that? The way behavioral economists and psychologists talk about that is at a proximate psychological level—saying things like, “Well, it feels good to cooperate with other people. You care about others and that’s why you’re willing to pay costs to help them. You have social preferences.” …
Most people, both in the scientific world and among laypeople, are of the former opinion, which is that we are by default selfish—we have to use rational deliberation to make ourselves do the right thing. I try to think about this question from a theoretical principle position and say, what should it be? From a perspective of either evolution or strategic reasoning, which of these two stories makes more sense, and should we expect to observe?
If you think about it that way, the key question is “where do our intuitive defaults come from?” There’s all this work in behavioral economics and psychology on heuristics and biases which suggests that these intuitions are usually rules of thumb for the behavior that typically works well. It makes sense: If you’re going to have something as your default, what should you choose as your default? You should choose the thing that works well in general. In any particular situation, you might stop and ask, “Does my rule of thumb fit this specific situation?” If not, then you can override it….(More)”
Innovation Labs: 10 Defining Features
Essay by Lidia Gryszkiewicz, Tuukka Toivonen, & Ioanna Lykourentzou: “Innovation labs, with their aspirations to foster systemic change, have become a mainstay of the social innovation scene. Used by city administrations, NGOs, think tanks, and multinational corporations, labs are becoming an almost default framework for collaborative innovation. They have resonance in and across myriad fields: London’s pioneering Finance Innovation Lab, for example, aims to create a system of finance that benefits “people and planet”; the American eLab is searching for the future of the electricity sector; and the Danish MindLab helps the government co-create better social services and solutions. Hundreds of other, similar initiatives around the world are taking on a range of grand challenges (or, if you prefer, wicked problems) and driving collective social impact.
Yet for all their seeming popularity, labs face a basic problem that closely parallels the predicament of hub organizations: There is little clarity on their core features, and few shared definitions exist that would make sense amid their diverse themes and settings. …
Building on observations previously made in the SSIR and elsewhere, we contribute to the task of clarifying the logic of modern innovation labs by distilling 10 defining features. …
1. Imposed but open-ended innovation themes…
2. Preoccupation with large innovation challenges…
3. Expectation of breakthrough solutions…
4. Heterogeneous participants…
5. Targeted collaboration…
6. Long-term perspectives…
7. Rich innovation toolbox…
8. Applied orientation…
9. Focus on experimentation…
10. Systemic thinking…
In a recent academic working paper, we condense the above into this definition: An innovation lab is a semi-autonomous organization that engages diverse participants—on a long-term basis—in open collaboration for the purpose of creating, elaborating, and prototyping radical solutions to pre-identified systemic challenges…(More)”
Wikipedia’s not as biased as you might think
Ananya Bhattacharya in Quartz: “The internet is as open as people make it. Often, people limit their Facebook and Twitter circles to likeminded people and only follow certain subreddits, blogs, and news sites, creating an echo chamber of sorts. In a sea of biased content, Wikipedia is one of the few online outlets that strives for neutrality. After 15 years in operation, it’s starting to see results
Researchers at Harvard Business School evaluated almost 4,000 articles in Wikipedia’s online database against the same entries in Encyclopedia Brittanica to compare their biases. They focused on English-language articles about US politics, especially controversial topics, that appeared in both outlets in 2012.
“That is just not a recipe for coming to a conclusion,” Shane Greenstein, one of the study’s authors, said in an interview. “We were surprised that Wikipedia had not failed, had not fallen apart in the last several years.”
Greenstein and his co-author Feng Zhu categorized each article as “blue” or “red.” Drawing from research in political science, they identified terms that are idiosyncratic to each party. For instance, political scientists have identified that Democrats were more likely to use phrases such as “war in Iraq,” “civil rights,” and “trade deficit,” while Republicans used phrases such as “economic growth,” “illegal immigration,” and “border security.”…
“In comparison to expert-based knowledge, collective intelligence does not aggravate the bias of online content when articles are substantially revised,” the authors wrote in the paper. “This is consistent with a best-case scenario in which contributors with different ideologies appear to engage in fruitful online conversations with each other, in contrast to findings from offline settings.”
More surprisingly, the authors found that the 2.8 million registered volunteer editors who were reviewing the articles also became less biased over time. “You can ask questions like ‘do editors with red tendencies tend to go to red articles or blue articles?’” Greenstein said. “You find a prevalence of opposites attract, and that was striking.” The researchers even identified the political stance for a number of anonymous editors based on their IP locations, and the trend held steadfast….(More)”
Learning Privacy Expectations by Crowdsourcing Contextual Informational Norms
“The advent of social apps, smart phones and ubiquitous computing has brought a great transformation to our day-to-day life. The incredible pace with which the new and disruptive services continue to emerge challenges our perception of privacy. To keep apace with this rapidly evolving cyber reality, we need to devise agile methods and frameworks for developing privacy-preserving systems that align with evolving user’s privacy expectations.
Previous efforts have tackled this with the assumption that privacy norms are provided through existing sources such law, privacy regulations and legal precedents. They have focused on formally expressing privacy norms and devising a corresponding logic to enable automatic inconsistency checks and efficient enforcement of the logic.
However, because many of the existing regulations and privacy handbooks were enacted well before the Internet revolution took place, they often lag behind and do not adequately reflect the application of logic in modern systems. For example, the Family Rights and Privacy Act (FERPA) was enacted in 1974, long before Facebook, Google and many other online applications were used in an educational context. More recent legislation faces similar challenges as novel services introduce new ways to exchange information, and consequently shape new, unconsidered information flows that can change our collective perception of privacy.
Crowdsourcing Contextual Privacy Norms
Armed with the theory of Contextual Integrity (CI) in our work, we are exploring ways to uncover societal norms by leveraging the advances in crowdsourcing technology.
In our recent paper, we present the methodology that we believe can be used to extract a societal notion of privacy expectations. The results can be used to fine tune the existing privacy guidelines as well as get a better perspective on the users’ expectations of privacy.
CI defines privacy as collection of norms (privacy rules) that reflect appropriate information flows between different actors. Norms capture who shares what, with whom, in what role, and under which conditions. For example, while you are comfortable sharing your medical information with your doctor, you might be less inclined to do so with your colleagues.
We use CI as a proxy to reason about privacy in the digital world and a gateway to understanding how people perceive privacy in a systematic way. Crowdsourcing is a great tool for this method. We are able to ask hundreds of people how they feel about a particular information flow, and then we can capture their input and map it directly onto the CI parameters. We used a simple template to write Yes-or-No questions to ask our crowdsourcing participants:
“Is it acceptable for the [sender] to share the [subject’s] [attribute] with [recipient] [transmission principle]?”
For example:
“Is it acceptable for the student’s professor to share the student’s record of attendance with the department chair if the student is performing poorly? ”
In our experiments, we leveraged Amazon’s Mechanical Turk (AMT) to ask 450 turkers over 1400 such questions. Each question represents a specific contextual information flow that users can approve, disapprove or mark under the Doesn’t Make Sense category; the last category could be used when 1) the sender is unlikely to have the information, 2) the receiver would already have the information, or 3) the question is ambiguous….(More)”
Predicting judicial decisions of the European Court of Human Rights: a Natural Language Processing perspective
Nikolaos Aletras et al at Peer J. Computer Science: “Recent advances in Natural Language Processing and Machine Learning provide us with the tools to build predictive models that can be used to unveil patterns driving judicial decisions. This can be useful, for both lawyers and judges, as an assisting tool to rapidly identify cases and extract patterns which lead to certain decisions. This paper presents the first systematic study on predicting the outcome of cases tried by the European Court of Human Rights based solely on textual content. We formulate a binary classification task where the input of our classifiers is the textual content extracted from a case and the target output is the actual judgment as to whether there has been a violation of an article of the convention of human rights. Textual information is represented using contiguous word sequences, i.e., N-grams, and topics. Our models can predict the court’s decisions with a strong accuracy (79% on average). Our empirical analysis indicates that the formal facts of a case are the most important predictive factor. This is consistent with the theory of legal realism suggesting that judicial decision-making is significantly affected by the stimulus of the facts. We also observe that the topical content of a case is another important feature in this classification task and explore this relationship further by conducting a qualitative analysis….(More)”
Civic Crowd Analytics: Making sense of crowdsourced civic input with big data tools
Paper by Tanja Aitamurto that: “… examines the impact of crowdsourcing on a policymaking process by using a novel data analytics tool called Civic CrowdAnalytics, applying Natural Language Processing (NLP) methods such as concept extraction, word association and sentiment analysis. By drawing on data from a crowdsourced urban planning process in the City of Palo Alto in California, we examine the influence of civic input on the city’s Comprehensive City Plan update. The findings show that the impact of citizens’ voices depends on the volume and the tone of their demands. A higher demand with a stronger tone results in more policy changes. We also found an interesting and unexpected result: the city government in Palo Alto mirrors more or less the online crowd’s voice while citizen representatives rather filter than mirror the crowd’s will. While NLP methods show promise in making the analysis of the crowdsourced input more efficient, there are several issues. The accuracy rates should be improved. Furthermore, there is still considerable amount of human work in training the algorithm….(More)”