This text is forwarded from WHO EPI-WIN, with a comment frm me below.
--
Infodemic Management News Flash
Thursday, 23 May 2024 | Issue #71
Discussions of artificial intelligence inevitably bring up discussions of existential reflection and ethical dilemmas. But defining more concretely the role of AI in health care, and health emergencies, remains nebulous and vaguely defined. It is a broadly defined research area with open questions to its real and imagined uses. Discussions around AI also inevitably raise questions such as, ‘Will AI take my job?’.
Any new technological leap is bound to invoke such discussion. But there are a number of factors that indicate that healthcare may be less at risk than other sectors. Critical thinking is necessary to much of healthcare, where one must be ready to weigh up benefits and risks of different approaches and solutions. Large language models can assist critical thinking, and can be directly useful in aiding infodemic managers distill large volumes of research into themes and narratives. However, this is different from actually doing the thinking involved in translating intelligence, into insights and recommendations.
The NHS Long Term Workforce Plan for example, gives an optimistic vision of the implementation of AI in diagnosis, treatment and disease management. And recent WHO publications note the potential of AI to identify harmful misinformation and "outdated" scientific information. However, in terms of addressing the needs of our modern information environment now, the operational use of AI in public remains grey. How would it function as part of an infodemic managers toolkit on a day-to-day basis?
LLMs such as ChatGPT, and generative tools like Dall-E or midjourney can be very useful content generating assistants for an infodemic manager whether that is as an analytical aid, data gathering on specific questions, language translation, creative ideation or artwork (see our newsflash cover art for example). For infodemic managers, the rapid development of generative AI tools may represent exciting options for evolution in the day-to-day nature of work. This may manifest as faster workflows, boosting workforce capacity, more rapid analysis, which may be especially useful in response interventions during acute emergencies. Some institutes such as the University of Zurich together with the WHO EURO region are conducting research specifically asking this question as to what the operational use of AI may look like in risk communication and infodemic management. How are you as an infodemic manager incorporating AI into your own workflow?
--
COMMENT (NPW): In addition to the above, and arguably most important, is the potential for AI to directly leapfrog other efforts to improve the availability and use of reliable healthcare information, whether for self-care or for patient care. This would eclipse any role to 'identify harmful misinformation and "outdated" scientific information'. The best way to tackle misinformation is to ensure the full availability and use of reliable healthcare information. With provisos, including and especially the integrity and trustworthiness of the managers of any given AI system, such systems have the potential to radically transform care. In my view, WHO should be taking the lead in champioing universal access to reliable healthcare information, and convening all stakeholders to develop a global strategy. This was also the resounding conclusion of our global survey. Our report is imminent.
HIFA profile: Neil Pakenham-Walsh is coordinator of HIFA (Healthcare Information For All), a global health community that brings all stakeholders together around the shared goal of universal access to reliable healthcare information. HIFA has 20,000 members in 180 countries, interacting in four languages and representing all parts of the global evidence ecosystem. HIFA is administered by Global Healthcare Information Network, a UK-based nonprofit in official relations with the World Health Organization. Email: neil@hifa.org
