Translated data: Microsoft has recently succeeded in transforming GPT-4 into a medical expert through prompt engineering, achieving an accuracy rate exceeding 90% in the medical field for the first time. Various prompting strategies of Medprompt have enabled GPT-4 to perform exceptionally well in the medical domain, with automatic generation of thought chains making the most significant contribution to performance improvement. The research results have garnered widespread attention in the industry, considered a technology with the potential to reshape the industry landscape. For detailed information, please refer to the paper link.