Important Design Questions for Algorithmic Ethics Consultation.

Abstract

Answering the design questions inherent to building and deploying machine learning tools (ML)—based on algorithms that can learn from and make predictions on large data sets without being explicitly programmed—requires explicit or implicit negotiation of ethical considerations (Char 2020). In applying ML tools to clinical ethics consultation, two important design questions, with ethical implications, stand out as needing discussion. First, why apply these tools to ethics consultation (i.e. why this target or clinical outcome?)? Second, what are likely impacts on the ethics workflow, including impacts on the professional identity of ethicists from introducing ML tools into this work?