At the heart of Veo 3 lies a complex web of generative models that have been trained on millions of video clips and image frames. When a user types a prompt, the system analyzes the language to understand the setting, tone, motion, and atmosphere. It then uses deep learning to construct a sequence of frames that plays like a real video.

YouTube is bringing its Veo 3 AI video generation model, which my colleague Allison Johnson recently called “a slop monger’s dream,” to YouTube Shorts later this summer, according to a Cannes Lions keynote from YouTube CEO Neal Mohan.