Suresh Venkatasubramanian of the University of Utah presented a method for finding disparate impact in algorithms last year at the ACM Conference on Knowledge Discovery and Data Mining. If you have ever searched for hotel rooms online, you have probably had this experience: surf over to another website to read a news story and the page fills up with ads for travel sites, offering deals on hotel rooms in the city you plan to visit. Buy something on Amazon, and ads for similar products will follow you around the Web. The practice of profiling people online means companies get more value from their advertising dollars and users are more likely to see ads that interest them. The practice has a downside, though, when the profiling is based on sensitive attributes, such as race, sex, or sexual orientation.

Traditional theories of social models conceptualize teams as distributed processors, disregarding the interdependence necessary to multi-task. Yet, interdependence characterizes social behavior. Instead, traditional theory favor cooperation, a state of least entropy production (LEP), without understanding the causes, limits or consequences of cooperation. As a simple example of interdependence, foraging prey overgraze forests free of predators. In our model, interdependence creates uncertainty, tradeoffs and signals (e.g., prices, coordination, innovation). Unlike individuals, the ability of teams to multitask reflects a quantum-like entanglement that represents maximum entropy production (MEP) when solving the problems signaled by society to improve its welfare. Our model supports findings that evolution in nature is driven by the MEP from making intelligent choices. Exploiting interdependence improves team intelligence, improves performance and reduces the risk of human error; forced cooperation disorganizes it by increasing the risk of error; e.g., if team cooperation improves teamwork, widespread forced cooperation in an autocracy or bureaucracy reduces social intelligence by adding unnecessary noise to signals. In our model, competition between teams self-organizes outsiders willing to sort through the noise for signals of the choices that improve social welfare (e.g., teams in courtrooms; science; entertainment; sports; businesses). Social systems organized around competition (e.g., stronger signals from robust checks and balances) better control a society by more correctly sizing teams to solve problems with fewer errors compared to autocracies or bureaucracies. Overall, we predict, the density of MEP directed at solving problems in a society with the constraints imposed from strong checks and balances, yet able to freely self-organize its labor and capital within those constraints, is denser.

As autonomous systems are becoming more and more pervasive, they often have to make decisions concerning moral and ethical values. There are many approaches to incorporating moral values in autonomous decision-making that are based on some sort of logical deduction. However, we argue here, in order for decision-making to seem persuasive to humans, it needs to reflect human values and judgments. Employing some insights from our ongoing researchusing features of the blackboard architecture for a context-aware recommender system, and a legal decision-making system that incorporates supra-legal aspects, we aim to explore if this architecture can also be adapted to implement a moral decision-making system that generates rationales that are persuasive to humans. Our vision is that such a system can be used as an advisory system to consider a situation from different moral perspectives, and generate ethical pros and cons of taking a particular course of action in a given context.

Rules offer a convenient means of limiting the operational scope of our AI programs so as to not transgress predictable moral boundaries. Yet the imposition of an operational morality based on mere rules will not turn our machines into moral agents, just the unthinking tools of moral designers. If we are to imbue our machines with a profound functional morality, we must first gift them with a moral imagination, for empathic morality — where one agent treats another as it would want to be treated itself — requires an ability to project oneself into the realms of the counterfactual. In this paper we thus explore the role of the moral imagination in generating new and inspiring stories. The creation of novel tales with a built-in moral requires that an artificial system possess the ability to guess at the morality of characters and their actions in novel settings and events. Our moralizing tale-spinner — which generates Aesop-style tales about human-like animals with identifiable human qualities — also faces another challenge: it must render these tales as micro-texts that can be distributed as tweets. As we shall also use metaphor to lend elasticity to our moral conceptions, these short stories, rich in animal metaphors, will comprise part of the daily output of the @MetaphorMagnet Twitterbot.

This paper describes and frames a central ethical issue–the liability problem–facing the regulation of artificial computational agents, including artificial intelligence (AI) and robotic systems, as they become increasingly autonomous, and supersede current capabilities. While it frames the issue in legal terms of liability and culpability, these terms are deeply imbued and interconnected with their ethical and moral correlate–responsibility. In order for society to benefit from advances in AI technology, it will be necessary to develop regulatory policies which manage the risk and liability of deploying systems with increasingly autonomous capabilities. However, current approaches to liability have difficulties when it comes to dealing with autonomous artificial agents because their behavior may be unpredictable to those who create and deploy them, and they will not be proper legal or moral agents. This problem is the motivation for a research project that will explore the fundamental concepts of autonomy, agency and liability; clarify the different varieties of agency that artificial systems might realize, including causal, legal and moral; and the illuminate the relationships between these. The paper will frame the problem of liability in autonomous agents, sketch out its relation to fundamental concepts in human legal and moral agency–including autonomy, agency, causation, intention, responsibility and culpability–and their applicability or inapplicability to autonomous artificial agents.

Here ought does require able--computationally and indeed logically intractable systems The question of Robot Ethics is difficult to resolve not because such as Asimov's laws are excluded (Myers, 2010). of the nature of Robots but because of the nature of What makes moral reasoning about intelligent artefacts Ethics. As with all normative considerations, robot ethics requires different from moral reasoning about natural entities is that that we decide what "really" matters--our most fundamental our obligations can be met not only through constructing the priorities. Are we more obliged to our biological socio-ethical system but also through specifications of the kin or to those with whom we share ideas? Do we value the artefacts. This is the definition of an artefact.

In practice, there are often explicit constraints on what representations or decisions are acceptable in an application of machine learning. For example it may be a legal requirement that a decision must not favour a particular group. Alternatively it can be that that representation of data must not have identifying information. We address these two related issues by learning flexible representations that minimize the capability of an adversarial critic. This adversary is trying to predict the relevant sensitive variable from the representation, and so minimizing the performance of the adversary ensures there is little or no information in the representation about the sensitive variable. We demonstrate this adversarial approach on two problems: making decisions free from discrimination and removing private information from images. We formulate the adversarial model as a minimax problem, and optimize that minimax objective using a stochastic gradient alternate min-max optimizer. We demonstrate the ability to provide discriminant free representations for standard test problems, and compare with previous state of the art methods for fairness, showing statistically significant improvement across most cases. The flexibility of this method is shown via a novel problem: removing annotations from images, from unaligned training examples of annotated and unannotated images, and with no a priori knowledge of the form of annotation provided to the model.

Clustering is an essential data mining tool that aims to discover inherent cluster structure in data. As such, the study of clusterability, which evaluates whether data possesses such structure, is an integral part of cluster analysis. Yet, despite their central role in the theory and application of clustering, current notions of clusterability fall short in two crucial aspects that render them impractical; most are computationally infeasible and others fail to classify the structure of real datasets. In this paper, we propose a novel approach to clusterability evaluation that is both computationally efficient and successfully captures the structure of real data. Our method applies multimodality tests to the (one-dimensional) set of pairwise distances based on the original, potentially high-dimensional data. We present extensive analyses of our approach for both the Dip and Silverman multimodality tests on real data as well as 17,000 simulations, demonstrating the success of our approach as the first practical notion of clusterability.

Computer Science Division, University of California, Berkeley, CA 94720 Dept. of Physics & MIT Kavli Institute, Massachusetts Institute of Technology, Cambridge, MA 02139 and Future of Humanity Institute, Oxford University, 16-17 St. Ebbe's str., Oxford OX1 1PT, UK Success in the quest for artificial intelligence has the potential to bring unprecedented benefits to humanity, and it is therefore worthwhile to investigate how to maximize these benefits while avoiding potential pitfalls. This article gives numerous examples (which should by no means be construed as an exhaustive list) of such worthwhile research aimed at ensuring that AI remains robust and beneficial. Artificial intelligence (AI) research has explored a variety of problems and approaches since its inception, but for the last 20 years or so has been focused on the problems surrounding the construction of intelligent agents - systems that perceive and act in some environment. In this context, the criterion for intelligence is related to statistical and economic notions of rationality-colloquially, the ability to make good decisions, plans, or inferences. The adoption of probabilistic representations and statistical learning methods has led to a large degree of integration and cross-fertilization between AI, machine learning, statistics, control theory, neuroscience, and other fields. The establishment of shared theoretical frameworks, combined with the availability of data and processing power, has yielded remarkable successes in various component tasks such as speech recognition, image classification, autonomous vehicles, machine translation, legged locomotion, and question-answering systems. As capabilities in these areas and others cross the threshold from laboratory research to economically valuable technologies, a virtuous cycle takes hold whereby even small improvements in performance are worth large sums of money, prompting greater investments in research. There is now a broad consensus that AI research is progressing steadily, and that its impact on society is likely to increase.

The need for new methods to deal with big data is a common theme in most scientific fields, although its definition tends to vary with the context. Statistical ideas are an essential part of this, and as a partial response, a thematic program on statistical inference, learning, and models in big data was held in 2015 in Canada, under the general direction of the Canadian Statistical Sciences Institute, with major funding from, and most activities located at, the Fields Institute for Research in Mathematical Sciences. This paper gives an overview of the topics covered, describing challenges and strategies that seem common to many different areas of application, and including some examples of applications to make these challenges and strategies more concrete.