Doctor's prescriptions, a cryptic mix of abbreviations, brand names, and local jargon, are notoriously difficult to decipher. But what if AI could help translate this medical shorthand into clear, understandable instructions? Researchers explored this possibility using ChatGPT, the popular large language model, to tackle the challenge of deciphering and expanding medication prescriptions. Focusing on Thai medical records, which often blend English and Thai medical terms with unique abbreviations, the study aimed to see if ChatGPT could bring clarity to these complex instructions. The researchers used a combination of "zero-shot" and "few-shot" learning approaches, providing ChatGPT with either no examples or a limited set of correctly annotated prescriptions to guide its learning. They then tested how well ChatGPT could identify key entities like medication names, dosages, units, and instructions. The results were impressive. With just a few examples to learn from, ChatGPT achieved a high accuracy in identifying and extracting information from the prescriptions. This suggests that even with minimal training data, large language models can become proficient in deciphering medical shorthand. Interestingly, the model struggled most with units of measurement, which were often omitted in the original prescriptions. However, the bigger concern for medication information is getting it precisely correct, and ChatGPT remarkably avoided "hallucinating" or fabricating information. It preferred to not expand unknown elements rather than create potentially harmful misinformation. This cautious approach is crucial for patient safety. The study's findings have significant real-world implications. Automating the interpretation of prescriptions could reduce errors, improve communication between healthcare providers and patients, and even facilitate cross-lingual understanding of medical instructions. Imagine a world where language barriers in medicine become a thing of the past, thanks to AI-powered translation tools. While more research is needed, particularly with larger datasets and different languages, this study demonstrates the potential of LLMs like ChatGPT to revolutionize how we process and understand medical information. It opens doors to a future where AI can enhance clarity, accuracy, and safety in healthcare, one prescription at a time.