1. Introduction to Generative AI
Generative AI (GAI) refers to algorithms (often using large language models, or LLMs) that can produce new content—like text, images, or even synthetic data—based on patterns learned from vast training data. In simple terms, it’s like an extremely advanced predictive text system: you give it a prompt or question, and it generates a human-like response by predicting what should come next, informed by everything it has learned. Unlike traditional AI that might just classify or flag data, GAI creates new sentences, explanations, or ideas that didn’t exist before (albeit based on what it has seen).
How it works
Imagine having a very well-read assistant in your computer. You ask a question, and this assistant has read millions of books, articles, and webpages. It doesn’t memorize them word-for-word, but it has learned the patterns of language and knowledge. When you ask it something, it tries to predict a useful answer by assembling words that often appear together in contexts similar to your question. The result can be a surprisingly coherent paragraph or explanation. For example, if you ask, “What are the symptoms of dengue fever?”, a GAI tool can generate a summary of dengue symptoms as if a doctor wrote it, because it has “seen” so many descriptions of dengue in its training.me text
How it works
GAI today Is No Longer Just "Text-Only" — It’s Multimodal Old GAI was like having a super-smart assistant that could only read and write. You typed in a question, and it replied with text. Now, GAI is becoming multimodal, which means it can see, hear, speak, and even understand pictures and videos. It's like your assistant is no longer just reading books — it's: – Looking at X-rays, scans, and photos 🩻📸 – Listening to what you say in any language 🎤🌏 – Talking back to you naturally 🔊 – Watching and describing video clips 🎬
Examples of GAI in Healthcare:
Clinical Decision Support Imagine a GP in Klang Valley seeing a patient with vague constitutional symptoms. Instead of going through multiple disconnected apps and notes, the doctor consults an integrated AI that: – Synthesizes current guidelines (e.g., NICE, UpToDate) – Reviews patient’s EMR – Flags concerning patterns like early hematologic malignancy or autoimmune disease – Suggests differentials and appropriate first-line investigations
Examples of GAI in Healthcare:
Ambient Documentation & Scribes Imagine a clinic where doctors no longer type while patients talk. Instead, smart glasses or ambient microphones record the consultation (with consent), and: – The GAI writes SOAP notes, referral letters, MDM summaries – Provides ICD-10 codes and CPT suggestions – Follows PDPA and encrypts data by default
Examples of GAI in Healthcare:
Decision support & information retrieval: Need a quick refresher on the latest management of a condition? Generative AI tools can summarize clinical guidelines or journal articles. They can’t replace clinical judgment, but they can augment your knowledge by quickly synthesizing information. Some advanced models have even shown ability in improving diagnostics or suggesting treatment plans by analyzing patient data (with proper privacy measures) though caution is needed (AI is not always correct).
Examples of GAI in Healthcare:
Patient education materials: Busy GPs can use GAI to create easy-to-understand explanations for patients. For instance, you can ask it to “Explain hypertension to a 50-year-old patient in simple language”, and it will produce a patient-friendly explanation that you can double-check and hand out. It’s like having a health brochure writer on call.
Examples of GAI in Healthcare:
Medical education: In teaching or self-learning, GAI can generate quiz questions, clinical scenarios, or even role-play as a virtual patient. For example, a GP teaching trainees can use AI to simulate a patient with symptoms and have the trainee practice asking it questions. Studies in medical education suggest GAI’s biggest potential is serving as a “virtual patient” or a tutor for clinical decision-making practice It can also help students by explaining complex topics in simpler terms, or generating mnemonics and study summaries.
SUMMARY