Date
Presentation at an online IILP Purdue Project webinar.
Abstract
The development of artificial intelligence (AI) techniques is rapidly changing the processing of and access to health information. Opportunities include faster and more user-friendly access to information, summarisation, translation and processing of data and information for diverse purposes and user groups including clinicians and patients. AI shows also considerable promise, for example, in providing support for decision-making in diagnostics and healthcare.
There are, however, multiple obstacles and open questions on the way to realising the envisioned opportunities. A key question is understanding how different AI techniques work in relation to health information and data, essentially developing an adequate level of literacy in the nexus of the two. Instead of merely developing a critical understanding the outputs generated by diverse fast evolving AI algorithms, the major question is to understand what AI does to health information, informing and getting informed about health. While the public debate on healthcare AI use might give an impression that the major issues relate to privacy and limited technical reliability of AI-based outputs, there are more fundamental epistemic and practical questions relating to who should be health information literate and how, how to measure and describe health information literacy, how to trust and whom, and who should have the responsibility.
The webinar draws from a long-line of information behaviour and literacy research in the nexus of health information and digital technologies, including work on AI and process documentation to make remarks on the implications of the use of AI techniques on health information practices and health information literacy. The presentation suggests that one of the most critical questions might be how and where to resist the ’complex easiness’ of AI, how to keep mastering health information as deeply complex and difficult issue that is navigable only with considerable effort, and how to integrate AI in facilitating relevant aspects of that still complex and demanding exercise.
- Log in to post comments
