The Hastings Center Report· 2019Q1
Artificial Intelligence and Black‐Box Medical Decisions: <i>Accuracy versus Explainability</i>
- 753citations
- Q1SCImago
- 2019year
Short summary
Opaque AI decisions in medicine are more common than critics assume, and their accuracy can be more important than explainability when causal knowledge is incomplete.
AI-generated from the title and abstract; the full text is not read.
Key points
- The most accurate machine learning techniques in medicine often function as 'black boxes,' lacking transparency in their decision-making processes.
- The ethical concerns regarding ceding medical decisions to opaque AI systems are based on an underestimation of how frequently opaque reasoning already occurs in clinical practice.
- When medical knowledge of causal systems is incomplete, the empirical accuracy of an AI's prediction or diagnosis can be ethically prioritized over its explainability.
AI-generated from the title and abstract; the full text is not read.
Abstract
Although decision-making algorithms are not new to medicine, the availability of vast stores of medical data, gains in computing power, and breakthroughs in machine learning are accelerating the pace of their development, expanding the range of questions they can address, and increasing their predictive power. In many cases, however, the most powerful machine learning techniques purchase diagnostic or predictive accuracy at the expense of our ability to access "the knowledge within the machine." Without an explanation in terms of reasons or a rationale for particular decisions in individual cases, some commentators regard ceding medical decision-making to black box systems as contravening the profound moral responsibilities of clinicians. I argue, however, that opaque decisions are more common in medicine than critics realize. Moreover, as Aristotle noted over two millennia ago, when our knowledge of causal systems is incomplete and precarious-as it often is in medicine-the ability to explain how results are produced can be less important than the ability to produce such results and empirically verify their accuracy.
The authors' abstract, as published at the source. The Hastings Center Report, 2019 · DOI ↗
Continue with a free account
Ask the paper: 3 free questions a day about this paper; save it, get its citation, new summaries every day for your field. Takeaways are Premium.
Continue free on the webSign in with Google or Apple; no card needed. You come back to this paper.
On your phone:
Field: Health Informatics
Health InformaticsMedicine