The danger of human operators devolving responsibility to machines and failing to detect cases where they fail has been recognised for many years by industrial psychologists and engineers studying the human operators of complex machines. We call it "the control problem", understood as the tendency of the human within a human-machine control loop to become complacent, over-reliant or unduly diffident when faced with the outputs of a reliable autonomous system. While the control problem has been investigated for some time, up to this point its manifestation in machine learning contexts has not received serious attention. This paper aims to fill that gap. We argue that, except in certain special circumstances, algorithmic decision tools should not be used in high-stakes or safety-critical decisions unless the systems concerned are significantly "better than human" in the relevant domain or subdomain of decision-making. More concretely, we recommend three strategies to address the control problem, the most promising of which involves a complementary (and potentially dynamic) coupling between highly proficient algorithmic tools and human agents working alongside one another. We also identify six key principles which all such human-machine systems should reflect in their design. These can serve as a framework both for assessing the viability of any such human-machine system as well as guiding the design and implementation of such systems generally.
The publication of the latest contribution to the alcohol-in-pregnancy debate, and the now customary flurry of media attention it generated, have precipitated the renewal of a series of ongoing debates about safe levels of consumption and responsible prenatal conduct. The University College London (UCL) study's finding that low levels of alcohol did not contribute to adverse behavioural outcomes-and may indeed have made a positive contribution in some cases-is unlikely to be the last word on the subject. Proving a negative correlation is notoriously difficult (technically, impossible), and other studies have offered alternative claims. The author is not an epidemiologist, and the purpose of this article is not to evaluate the competing empirical claims. However, the question of what information and advice healthcare practitioners ought to present to pregnant women, or prospectively or potentially pregnant women, in a situation of uncertainty is one to which healthcare ethicists may have a contribution to make. In this article, it is argued that the total abstinence policy advocated by the UK's Department of Health, and even more stridently by the British Medical Association, sits uneasily with recent data and is far from ethically unproblematic. In particular, the "precautionary" approach advocated by these bodies displays both scant regard for the autonomy of pregnant and prospectively pregnant women and a confused grasp of the principles of beneficence and non-maleficence.
The recent case of David Bradley, who shot and killed four members of his family after telling his doctor he 'wanted to kill someone', has raised the question of whether a healthcare professional could ever be held liable for failing to take steps to constrain a potentially dangerous patient. Until recently, it was considered that the United Kingdom courts would be reluctant to impose a duty to protect third parties. However, the European Court of Human Rights' decision in Osman v UK--while not directly concerning healthcare professionals--has opened the door for just such a duty. When this duty will arise, and how it can be discharged, remain challenging questions. Furthermore, healthcare professionals face the unenviable task of balancing competing duties, in which the rights--and safety--of their patients must also be borne in mind.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.