Decision Technologies in Medical Research and Practice: Practical Considerations, Ethical Implications, and the Need for Dialectic Evaluation

Abstract

The use of computer programs to generate hypotheses for basic science research has been the touchstone of recent and growing debate. These programs engage amounts of scientific data that are insurmountable for human cognitive processing, prompting questions of whether the capability for such computer-enabled hypothesis generation can and/or will fundamentally alter creativity in scientific research and discovery. The technology employed for hypothesis generation is part of a family of computer-based algorithms that confer putatively enhanced ability to discover, predict, and recommend novel and fruitful interrelations within and across various types and vast ranges of data. We refer to these technical tools as "decision technologies" because they simulate, automate, and integrate cognitive tasks via the employment of computational algorithms and forms of artificial intelligence to parse the most "beneficial" courses of action given an extensive diversity of potential options. As decision technologies become more relevant across a range of human enterprises, important questions that span technical and ethical domains arise. How can we delimit the power of these technologies while maintaining the autonomy and moral responsibility of the human decision makers? What are the responsibilities for incorrect or harmful outcomes fostered by these technologies, and who shall bear them? And how should ethical discourse proceed to effectively oversee and guide the development and use of these technologies in medical research, clinical practice, and more broadly in public life? Using "machine science" as the exemplar and starting point, additional technical and ethical issues are discussed here, and an ethically dialectic approach to assessing, directing and remaining prepared for the contingencies generated by decision technologies is presented.

Ähnliche Arbeiten