AI-powered systems for healthcare:
好处： fast； personalized； reduce costs
挑战： complex organizational changes； long-term social consequences
Sociotechnical challenges include the potential for biased algorithms that benefit certain subgroups of patient populations (e.g., using race to predict treatments in the presence of health disparities), effects on patient–clinician communication, the need for new skills and workflows to practice medicine, and perceptions and anxieties about people’s future role within medicine (e.g., whether image recognition systems perform better than radiologists).
因此，我们不仅要考虑算法等技术问题，也要考虑normative, regulatory, and ethical challenges 以及其他 potential negative effects of AI on society
本文还给出了一些讨论的问题：
人机关系； 人对机器的信任问题
| 文献题目 | 去谷歌学术搜索 | ||||||||||
| Identifying Challenges and Opportunities in Human–AI Collaboration in Healthcare | |||||||||||
| 文献作者 | |||||||||||
| 文献发表年限 | 2019 | ||||||||||
| 文献关键字 | |||||||||||
| CSCW; Healthcare | |||||||||||
| 摘要描述 | |||||||||||
| The proposed workshop will identify research questions that will enable the field to uncover the types of work, labor relations, and social impacts that should be considered when designing AIbased healthcare technology. The workshop aims to outline key challenges, guidelines, and future agendas for the field, and provide collaboration opportunities for CSCW researchers, social scientists, AI researchers, clinicians, and relevant stakeholders in healthcare, to share their perspectives and co-create sociotechnical approaches to tackle timely issues related to AI and automation in healthcare work. | |||||||||||
