The goal of this work is to help mitigate the already existing gender wage gap by supplying unbiased job recommendations based on resumes from job seekers. We employ a generative adversarial network to remove gender bias from word2vec representations of 12M job vacancy texts and 900k resumes. Our results show that representations created from recruitment texts contain algorithmic bias and that this bias results in real-world consequences for recommendation systems. Without controlling for bias, women are recommended jobs with significantly lower salary in our data. With adversarially fair representations, this wage gap disappears, meaning that our debiased job recommendations reduce wage discrimination. We conclude that adversarial debiasing of word representations can increase real-world fairness of systems and thus may be part of the solution for creating fairness-aware recommendation systems.

Predictive models are increasingly used to make various consequential decisions in high-stakes domains such as healthcare, finance, and policy. It becomes critical to ensure that these models make accurate predictions, are robust to shifts in the data, do not rely on spurious features, and do not unduly discriminate against minority groups. To this end, several approaches spanning various areas such as explainability, fairness, and robustness have been proposed in recent literature. Such approaches need to be human-centered as they cater to the understanding of the models to their users. However, there is a research gap in understanding the human-centric needs and challenges of monitoring machine learning (ML) models once they are deployed. To fill this gap, we conducted an interview study with 13 practitioners who have experience at the intersection of deploying ML models and engaging with customers spanning domains such as financial services, healthcare, hiring, online retail, computational advertising, and conversational assistants. We identified various human-centric challenges and requirements for model monitoring in real-world applications. Specifically, we found the need and the challenge for the model monitoring systems to clarify the impact of the monitoring observations on outcomes. Further, such insights must be actionable, robust, customizable for domain-specific use cases, and cognitively considerate to avoid information overload.

Integer sequences are of central importance to the modeling of concepts admitting complete finitary descriptions. We introduce a novel view on the learning of such concepts and lay down a set of benchmarking tasks aimed at conceptual understanding by machine learning models. These tasks indirectly assess model ability to abstract, and challenge them to reason both interpolatively and extrapolatively from the knowledge gained by observing representative examples. To further aid research in knowledge representation and reasoning, we present FACT, the Finitary Abstraction Comprehension Toolkit.

The forecasting of credit default risk has been an active research field for several decades. Historically, logistic regression has been used as a major tool due to its compliance with regulatory requirements: transparency, explainability, and fairness. In recent years, researchers have increasingly used complex and advanced machine learning methods to improve prediction accuracy. Even though a machine learning method could potentially improve the model accuracy, it complicates simple logistic regression, deteriorates explainability, and often violates fairness. In the absence of compliance with regulatory requirements, even highly accurate machine learning methods are unlikely to be accepted by companies for credit scoring. In this paper, we introduce a novel class of monotonic neural additive models, which meet regulatory requirements by simplifying neural network architecture and enforcing monotonicity. By utilizing the special architectural features of the neural additive model, the monotonic neural additive model penalizes monotonicity violations effectively. Consequently, the computational cost of training a monotonic neural additive model is similar to that of training a neural additive model, as a free lunch. We demonstrate through empirical results that our new model is as accurate as black-box fully-connected neural networks, providing a highly accurate and regulated machine learning method.

Artificial intelligence (AI) has the potential to greatly improve society, but as with any powerful technology, it comes with heightened risks and responsibilities. Current AI research lacks a systematic discussion of how to manage long-tail risks from AI systems, including speculative long-term risks. Keeping in mind the potential benefits of AI, there is some concern that building ever more intelligent and powerful AI systems could eventually result in systems that are more powerful than us; some say this is like playing with fire and speculate that this could create existential risks (x-risks). To add precision and ground these discussions, we provide a guide for how to analyze AI x-risk, which consists of three parts: First, we review how systems can be made safer today, drawing on time-tested concepts from hazard analysis and systems safety that have been designed to steer large processes in safer directions. Next, we discuss strategies for having long-term impacts on the safety of future systems. Finally, we discuss a crucial concept in making AI systems safer by improving the balance between safety and general capabilities. We hope this document and the presented concepts and tools serve as a useful guide for understanding how to analyze AI x-risk.

Human relationships are complex processes that often involve following certain rules that regulate interactions and/or expected outcomes. These rules may be imposed by an authority or established by society. In multi-agent systems, normative systems have extensively addressed aspects such as norm synthesis, norm conflict detection, as well as norm emergence. However, if human behaviour is to be adequately simulated, not only normative aspects but also emotional aspects have to be taken into account. In this paper, we propose a Jason agent architecture that incorporates norms and emotions in its reasoning process to determine which plan (actions) to execute. The proposal is evaluated through a scenario based on a social network, which allows us to analyse the benefits of using emotional normative agents to achieve simulations closer to real human world.

A challenging task when generating summaries of legal documents is the ability to address their argumentative nature. We introduce a simple technique to capture the argumentative structure of legal documents by integrating argument role labeling into the summarization process. Experiments with pretrained language models show that our proposed Figure 1: Overview of our approach.

Recent years have seen a surge of interest in the field of explainable AI (XAI), with a plethora of algorithms proposed in the literature. However, a lack of consensus on how to evaluate XAI hinders the advancement of the field. We highlight that XAI is not a monolithic set of technologies -- researchers and practitioners have begun to leverage XAI algorithms to build XAI systems that serve different usage contexts, such as model debugging and decision-support. Algorithmic research of XAI, however, often does not account for these diverse downstream usage contexts, resulting in limited effectiveness or even unintended consequences for actual users, as well as difficulties for practitioners to make technical choices. We argue that one way to close the gap is to develop evaluation methods that account for different user requirements in these usage contexts. Towards this goal, we introduce a perspective of contextualized XAI evaluation by considering the relative importance of XAI evaluation criteria for prototypical usage contexts of XAI. To explore the context dependency of XAI evaluation criteria, we conduct two survey studies, one with XAI topical experts and another with crowd workers. Our results urge for responsible AI research with usage-informed evaluation practices, and provide a nuanced understanding of user requirements for XAI in different usage contexts.

Interview In 2008, David Holz co-founded a hardware peripheral firm called Leap Motion. He ran it until last year when he left to create Midjourey. Midjourney in its present form is a social network for creating AI-generated art from a text prompt – type a word or phrase at the input prompt and you'll receive an interesting or perhaps wonderful image on screen after about a minute of computation. It's similar in some respects to OpenAI's DALL-E 2. Midjourney image of the sky and clouds, using the text prompt "All this useless beauty." Both are the result of large AI models trained on vast numbers of images. But Midjourney has its own distinctive style, as can be seen from this Twitter thread.

Abundant data, widespread digitisation, and attractive efficiency gains have driven the development and use of Artificial Intelligence (AI). However, this rapid growth is not isolated from human rights abuses which often stem from the way AI technologies are deployed. In July 2022, Aapti Institute, a Bengaluru-based tech think-tank, in collaboration with the Business and Human Rights (Asia) programme at UNDP India, examined the impact of AI deployment on the human rights of consumers in finance, healthcare, and on the labour force in gig work and retail in India. This work builds on existing research, such as the Human Rights Guide by the Danish Institute on Human Rights, which has found that a human rights-respecting approach by businesses can enhance individual and community well-being and drive sustainable economic growth. Our research identified numerous sector-specific risks and found commonalities across sectors.