Artificial intelligence for healthcare: restrained development despite impressive applications

Artificial intelligence (AI) remains poorly understood and its rapid growth raises concerns reminiscent of dystopian narratives. AI has shown the capability of producing new medical content and improving management through optimization and standardization, which shortens queues, while its complete reliance on technical solutions threatens the traditional doctor-patient bond.

Based on the World Economic Forum's emphasis on the need for faster AI adoption in the medical field, we highlight current gaps in the understanding of its application and offer a set of priorities for future research. The historic review of AI and the latest publications point at barriers like complexity and fragmented regulations, while assisted analysis of big data offers new insights. AI's potential in healthcare is linked to the breakthrough from rule-based computing, enabling autonomy through learning from experience and the capacity of reasoning. Without AI, protein folding would have remained unsolved, as emphasized by the Nobel-honored AlphaFold2 approach. It is expected that AI's role in diagnostics, disease control, geospatial health and epidemiology will lead to similar progress.

AI boosts efficiency, drives innovation, and solves complex problems but can also deepen biases and create security threats. Controlled progress requires industry collaboration leading to prompt acceleration of proper incorporation of AI into the health sphere. Cooperation between governments as well as both public and private sectors with a multi-actor approach is needed to effectively address these challenges. To fully harness AI's potential in accelerating healthcare reform and shorten queues, while maintaining the compassionate essence of healthcare, a well-coordinated approach involving all stakeholders is necessary.