OpenAI’s Sora in ophthalmology: revolutionary generative AI in eye health
Ethan Waisberg, Joshua Ong, Mouayad Masalkhi, Andrew G. Lee
Abstract
Sora is a state-of-the-art AI model developed by Open AI that has been engineered to generate realistic and imaginative scenes purely based on textual instructions [ 1 ]. This innovative application is a remarkable leap forward and highlights the advanced capabilities of modern AI in interpreting and visualizing complex narratives [ 1 ]. The technological foundation of Sora rests on Large Language Models (LLMs) and artificial video generation techniques. LLMs are advanced neural network architectures designed to understand, generate, and interpret human language in a highly sophisticated manner [ 2 , 3 , 4 , 5 ]. When combined with diffusion models for video generation, these AI systems can create detailed and dynamic visual content from text descriptions [ 1 ]. This involves processing the text to understand its meaning and context, and then translating them into a series of images that form a coherent video sequence.