This paper deals with the importance of developing codes of conduct for practitioners--be it journalists, doctors, attorneys, or other professions--that are encountering ethical issues when using computation, but do not have access to any framework of reference as to how to address those. At the same time, legal and technological developments are calling for establishing such guidelines, as shown in the European Union's and the United States' efforts in regulating a wide array of artificial intelligence systems, and in the resurgence of rule-based models through 'neurosymbolic' AI, a hybrid format that combines them with neural methods. Against this backdrop, we argue for taking a design-inspired approach when encoding professional ethics into a computational form, so as to co-create codes of conduct for computational practice across a wide range of fields.
翻译:本文论述了为遇到计算相关伦理问题但缺乏应对参考框架的从业者(如记者、医生、律师或其他职业)制定行为准则的重要性。与此同时,法律和技术发展也要求建立此类指导方针——欧盟与美国在广泛监管人工智能系统方面的努力,以及通过“神经符号”AI(一种结合符号方法与神经方法的混合模式)重新兴起的基于规则的模型,均印证了这一点。在此背景下,我们主张在将专业伦理编码为计算形式时采用设计导向的方法,从而为多领域的计算实践协同制定行为准则。