Policymakers around the world are increasingly considering how to prevent government uses of algorithms from producing injustices. One mechanism that has become a centerpiece of global efforts to regulate government algorithms is to require human oversight of algorithmic decisions. Despite the widespread turn to human oversight, these policies rest on an uninterrogated assumption: that people are able to oversee algorithmic decision-making. In this article, I survey 40 policies that prescribe human oversight of government algorithms and find that they suffer from two significant flaws. First, evidence suggests that people are unable to perform the desired oversight functions. Second, as a result of the first flaw, human oversight policies legitimize government uses of faulty and controversial algorithms without addressing the fundamental issues with these tools. Thus, rather than protect against the potential harms of algorithmic decision-making in government, human oversight policies provide a false sense of security in adopting algorithms and enable vendors and agencies to shirk accountability for algorithmic harms. In light of these flaws, I propose a more stringent approach for determining whether and how to incorporate algorithms into government decision-making. First, policymakers must critically consider whether it is appropriate to use an algorithm at all in a specific context. Second, before deploying an algorithm alongside human oversight, agencies or vendors must conduct preliminary evaluations of whether people can effectively oversee the algorithm.
相關內容
Humans' continual learning (CL) ability is closely related to Stability Versus Plasticity Dilemma that describes how humans achieve ongoing learning capacity and preservation for learned information. The notion of CL has always been present in artificial intelligence (AI) since its births. This paper proposes a comprehensive review of CL. Different from previous reviews that mainly focus on the catastrophic forgetting phenomenon in CL, this paper surveys CL from a more macroscopic perspective based on the Stability Versus Plasticity mechanism. Analogous to biological counterpart, "smart" AI agents are supposed to i) remember previously learned information (information retrospection); ii) infer on new information continuously (information prospection:); iii) transfer useful information (information transfer), to achieve high-level CL. According to the taxonomy, evaluation metrics, algorithms, applications as well as some open issues are then introduced. Our main contributions concern i) rechecking CL from the level of artificial general intelligence; ii) providing a detailed and extensive overview on CL topics; iii) presenting some novel ideas on the potential development of CL.
Governments must keep agricultural systems free of pests that threaten agricultural production and international trade. Biosecurity surveillance already makes use of a wide range of technologies, such as insect traps and lures, geographic information systems, and diagnostic biochemical tests. The rise of cheap and usable surveillance technologies such as remotely piloted aircraft systems (RPAS) presents value conflicts not addressed in international biosurveillance guidelines. The costs of keeping agriculture pest-free include privacy violations and reduced autonomy for farmers. We argue that physical and digital privacy in the age of ubiquitous aerial and ground surveillance is a natural right to allow people to function freely on their land. Surveillance methods must be co-created and justified through using ethically defensible processes such as discourse theory, value-centred design and responsible innovation to forge a cooperative social contract between diverse stakeholders. We propose an ethical framework for biosurveillance activities that balances the collective benefits for food security with individual privacy: (1) establish the boundaries of a biosurveillance social contract; (2) justify surveillance operations for the farmers, researchers, industry, the public and regulators; (3) give decision makers a reasonable measure of control over their personal and agricultural data; and (4) choose surveillance methodologies that give the appropriate information. The benefits of incorporating an ethical framework for responsible biosurveillance innovation include increased participation and accumulated trust over time. Long term trust and cooperation will support food security, producing higher quality data overall and mitigating against anticipated information gaps that may emerge due to disrespecting landholder rights
Results from randomized controlled trials (RCTs) help determine vaccination strategies and related public health policies. However, defining and identifying estimands that can guide policies in infectious disease settings is difficult, even in an RCT. The effects of vaccination critically depend on characteristics of the population of interest, such as the prevalence of infection, the number of vaccinated, and social behaviors. To mitigate the dependence on such characteristics, estimands (and study designs) that require conditioning or intervening on exposure to the infectious agent have been advocated. But a fundamental problem for both RCTs and observational studies is that exposure status is often unavailable or difficult to measure, which has made it impossible to apply existing methodology to study vaccine effects that account for exposure status. In this work, we present new results on this type of vaccine effects. Under plausible conditions, we show that point identification of certain relative effects is possible even when the exposure status is unknown. Furthermore, we derive sharp bounds on the corresponding absolute effects. We apply these results to estimate the effects of the ChAdOx1 nCoV-19 vaccine on SARS-CoV-2 disease (COVID-19) conditional on post-vaccine exposure to the virus, using data from a large RCT.
Recommender systems have been widely used in different application domains including energy-preservation, e-commerce, healthcare, social media, etc. Such applications require the analysis and mining of massive amounts of various types of user data, including demographics, preferences, social interactions, etc. in order to develop accurate and precise recommender systems. Such datasets often include sensitive information, yet most recommender systems are focusing on the models' accuracy and ignore issues related to security and the users' privacy. Despite the efforts to overcome these problems using different risk reduction techniques, none of them has been completely successful in ensuring cryptographic security and protection of the users' private information. To bridge this gap, the blockchain technology is presented as a promising strategy to promote security and privacy preservation in recommender systems, not only because of its security and privacy salient features, but also due to its resilience, adaptability, fault tolerance and trust characteristics. This paper presents a holistic review of blockchain-based recommender systems covering challenges, open issues and solutions. Accordingly, a well-designed taxonomy is introduced to describe the security and privacy challenges, overview existing frameworks and discuss their applications and benefits when using blockchain before indicating opportunities for future research.
The increasing concerns about data privacy and security drive an emerging field of studying privacy-preserving machine learning from isolated data sources, i.e., federated learning. A class of federated learning, \textit{vertical federated learning}, where different parties hold different features for common users, has a great potential of driving a great variety of business cooperation among enterprises in many fields. In machine learning, decision tree ensembles such as gradient boosting decision trees (GBDT) and random forest are widely applied powerful models with high interpretability and modeling efficiency. However, state-of-art vertical federated learning frameworks adapt anonymous features to avoid possible data breaches, makes the interpretability of the model compromised. To address this issue in the inference process, in this paper, we firstly make a problem analysis about the necessity of disclosure meanings of feature to Guest Party in vertical federated learning. Then we find the prediction result of a tree could be expressed as the intersection of results of sub-models of the tree held by all parties. With this key observation, we protect data privacy and allow the disclosure of feature meaning by concealing decision paths and adapt a communication-efficient secure computation method for inference outputs. The advantages of Fed-EINI will be demonstrated through both theoretical analysis and extensive numerical results. We improve the interpretability of the model by disclosing the meaning of features while ensuring efficiency and accuracy.
Online communities, like Wikipedia, produce valuable public information goods. Whereas some of these communities require would-be contributors to create accounts, many do not. Does this requirement catalyze cooperation or inhibit participation? Prior research provides divergent predictions but little causal evidence. We conduct an empirical test using longitudinal data from 136 natural experiments where would-be contributors to wikis were suddenly required to log in to contribute. Requiring accounts leads to a small increase in account creation, but reduces both high- and low-quality contributions from registered and unregistered participants. Although the change deters a large portion of low-quality participation, the vast majority of deterred contributions are of higher quality. We conclude that requiring accounts introduces an undertheorized tradeoff for public goods production in interactive communication systems.
Modern cars technologies are evolving quickly. They collect a variety of personal data and treat it on behalf of the car manufacturer to improve the drivers' experience. The precise terms of such a treatment are stated within the privacy policies accepted by the user when buying a car or through the infotainment system when it is first started. This paper uses a double lens to assess people's privacy while they drive a car. The first approach is objective and studies the readability of privacy policies that comes with cars. We analyse the privacy policies of twelve car brands and apply well-known readability indices to evaluate the extent to which privacy policies are comprehensible by all drivers. The second approach targets drivers' opinions to extrapolate their privacy concerns and trust perceptions. We design a questionnaire to collect the opinions of 88 participants and draw essential statistics about them. Our combined findings indicate that privacy is insufficiently understood at present as an issue deriving from driving a car, hence future technologies should be tailored to make people more aware of the issue and to enable them to express their preferences.
While human beings have a right to digital experiences that support, rather than diminish, their psychological wellbeing, technology designers lack research-based practices for ensuring psychological needs are met. To help address this gap, we draw on findings from over 30 years of research in psychology (specifically, self-determination theory) that has identified contextual factors shown to support psychological wellbeing. We translate these findings into a list of 15 heuristics and 30 design strategies to provide technology makers with theoretically grounded, research-based, and actionable ways to support wellbeing in user experience.
Fast developing artificial intelligence (AI) technology has enabled various applied systems deployed in the real world, impacting people's everyday lives. However, many current AI systems were found vulnerable to imperceptible attacks, biased against underrepresented groups, lacking in user privacy protection, etc., which not only degrades user experience but erodes the society's trust in all AI systems. In this review, we strive to provide AI practitioners a comprehensive guide towards building trustworthy AI systems. We first introduce the theoretical framework of important aspects of AI trustworthiness, including robustness, generalization, explainability, transparency, reproducibility, fairness, privacy preservation, alignment with human values, and accountability. We then survey leading approaches in these aspects in the industry. To unify the current fragmented approaches towards trustworthy AI, we propose a systematic approach that considers the entire lifecycle of AI systems, ranging from data acquisition to model development, to development and deployment, finally to continuous monitoring and governance. In this framework, we offer concrete action items to practitioners and societal stakeholders (e.g., researchers and regulators) to improve AI trustworthiness. Finally, we identify key opportunities and challenges in the future development of trustworthy AI systems, where we identify the need for paradigm shift towards comprehensive trustworthy AI systems.
BERT-based architectures currently give state-of-the-art performance on many NLP tasks, but little is known about the exact mechanisms that contribute to its success. In the current work, we focus on the interpretation of self-attention, which is one of the fundamental underlying components of BERT. Using a subset of GLUE tasks and a set of handcrafted features-of-interest, we propose the methodology and carry out a qualitative and quantitative analysis of the information encoded by the individual BERT's heads. Our findings suggest that there is a limited set of attention patterns that are repeated across different heads, indicating the overall model overparametrization. While different heads consistently use the same attention patterns, they have varying impact on performance across different tasks. We show that manually disabling attention in certain heads leads to a performance improvement over the regular fine-tuned BERT models.
注冊地址： 北京市海淀區羊坊店路18號2幢3層301-191