Adopting Generative AI with Precaution in Dentistry: A Review and Reflection

The progress in large language models (LLMs) brings much excitement and efforts in medical artificial intelligence, which could transform patient-doctor conversation while making joint medical decisions. LLMs, exemplified by ChatGPT, are proficient in grasping and generating text, and can perform tasks such as question answering, document summarising, and paraphrasing with a level of proficiency comparable to that of a human. Their potential applications span across various tasks in medicine, notably improving clinical patient care experience, advancing scientific medical research, and revolutionizing medical education. This survey critically examines the evolving landscape of medical large language models (Med LLMs), with a special focus on their application in stomatology. While Med LLMs are inevitably becoming an integral part to medical text processing and image processing, their use in enhancing clinical care requires extra precaution and assurance due to the stringent requirements on ethics and patient safety. The design, deployment and use of LLMs and services requires thorough risks analysis of technology misuse and potential harms. This survey looks into the current status, different prospects and challenges in LLMs development in medical use cases and ways to control and mitigates risks of generative artificial intelligence.