Argumentation-based dialogue systems, which can handle and exchange arguments through dialogue, have been widely researched. It is required that these systems have sufficient supporting information to argue their claims rationally; however, the systems often do not have enough of such information in realistic situations. One way to fill in the gap is acquiring such missing information from dialogue partners (information-seeking dialogue). Existing information-seeking dialogue systems are based on handcrafted dialogue strategies that exhaustively examine missing information. However, the proposed strategies are not specialized in collecting information for constructing rational arguments. Moreover, the number of system's inquiry candidates grows in accordance with the size of the argument set that the system deal with. In this paper, we formalize the process of information-seeking dialogue as Markov decision processes (MDPs) and apply deep reinforcement learning (DRL) for automatically optimizing a dialogue strategy. By utilizing DRL, our dialogue strategy can successfully minimize objective functions, the number of turns it takes for our system to collect necessary information in a dialogue. We conducted dialogue experiments using two datasets from different domains of argumentative dialogue. Experimental results show that the proposed formalization based on MDP works well, and the policy optimized by DRL outperformed existing heuristic dialogue strategies.

We consider how fair treatment in society for people with disabilities might be impacted by the rise in the use of artificial intelligence, and especially machine learning methods. We argue that fairness for people with disabilities is different to fairness for other protected attributes such as age, gender or race. One major difference is the extreme diversity of ways disabilities manifest, and people adapt. Secondly, disability information is highly sensitive and not always shared, precisely because of the potential for discrimination. Given these differences, we explore definitions of fairness and how well they work in the disability space. Finally, we suggest ways of approaching fairness for people with disabilities in AI applications.

Inc. shut down a project it had been developing for four years, a recruitment tool driven by machine learning. The concept was a simple and appealing one: at its core, the project aimed to develop an algorithm that would sort incoming job applications to isolate the short list for managers to use in making their final selections. Anyone who has been involved in such a process knows that isolating the top five or ten resumés from dozens of applicants is a time-consuming job. Any process that brings logic and speed to this stage of the recruitment chain can only be a good thing. But algorithms are only as good as the data used to drive them, and in the Amazon case the machine "learning" was based on patterns in applications submitted to the firm in the previous ten years.

Saudi Arabia made waves in late 2017 when it granted citizenship to a humanoid robot named Sophia, developed by the Hong Kong-based Hanson Robotics. What those rights technically include, and what the move might mean for other robots worldwide, remains unclear. But the robot itself wasted no time in taking advantage of her new, high profile to campaign for women's rights in her adopted country. This would be the same Sophia that, in a CNBC interview with her creator, Dr. David Hanson, said that she would "destroy all humans." So, granting legal rights to robots clearly remains a complicated subject, even if it is done primarily as a PR stunt to promote an IT conference, as was the case in Saudi Arabia.

In 2014, Amazon developed an artificial intelligence (AI) recruitment tool that began to discriminate against female job applicants. A year later, a user of Google Photos discovered that the program was labeling his black friends as gorillas. And in 2018 it emerged that an algorithm that analyzed the risk of recidivism by a million US defendants made as many mistakes as any human being with no training in criminal justice. Decisions that were once made by human beings are now being made by AI systems. Some programs deal with hiring, others with loan approvals, medical diagnoses and even court rulings.

Everyone wants to be safe. But paradoxically, sometimes the policies we implement to guarantee our safety end up making us much worse off than if we had done nothing at all. It is counterintuitive, but this is the well-established calculus of the world of risk analysis. When we consider the future of AI and the public policies that will shape its evolution, it is vital to keep that insight in mind. While AI-enabled technologies can pose some risks that should be taken seriously, it is important that public policy not freeze the development of life-enriching innovations in this space based on speculative fears of an uncertain future.

Group fairness is an important concern for machine learning researchers, developers, and regulators. However, the strictness to which models must be constrained to be considered fair is still under debate. The focus of this work is on constraining the expected outcome of subpopulations in kernel regression and, in particular, decision tree regression, with application to random forests, boosted trees and other ensemble models. While individual constraints were previously addressed, this work addresses concerns about incorporating multiple constraints simultaneously. The proposed solution does not affect the order of computational or memory complexity of the decision trees and is easily integrated into models post training.

Artificial intelligence is "shockingly" racist and sexist, a study has revealed. Researchers looked at a range of systems and datasets and found examples where AI had provided inaccurate information for women and minorities. In one example, the team from Massachusetts Institute of Technology looked at an income prediction system and discovered it was twice as likely to misclassify female employees as low-income and male employee as high-income. However, the team was able to adjust the system to make sure it was less biased. When researchers increased the dataset by a factor of 10, they found the mistakes decreased by 40 percent.

The voice that responds when you say, "OK Google." These virtual assistants rely on artificial intelligence. They are increasingly ubiquitous, and they are female. Kate Devlin, a technology expert and senior lecturer at King's College, London, says it may stem from biases that can lurk deep in human thought, perhaps even unnoticed. She recounts how, when she asked a developer of one of the digital-assistants why he chose a female voice, his answer was, "I didn't really think about it."

Machine learning algorithms process vast quantities of data and spot correlations, trends and anomalies, at levels far beyond even the brightest human mind. But as human intelligence relies on accurate information, so too do machines. Algorithms need training data to learn from. This training data is created, selected, collated and annotated by humans. And therein lies the problem.