From a pinhole camera sketched in 1544 to machines that can recognize faces, read X-rays, and navigate self-driving cars — the story of computer vision spans nearly five centuries of human curiosity and computational ambition. This video traces the remarkable journey of how we taught machines to see. It begins with Gemma Frisius and his groundbreaking illustration of the camera obscura, which first captured the physics of how light forms an image — a principle that modern sensors still rely on today. From there, it leaps to 1963, where MIT's Larry Roberts, often called the father of computer vision, developed a system that could digitize 2D photographs and reconstruct them as 3D geometric shapes like cubes and wedges, pioneering the edge detection techniques that still underpin modern machine perception. Then comes one of AI history's most instructive failures: the 1966 MIT Summer Vision Project, where researchers boldly believed they could solve machine vision in a single summer. They couldn't. That failure didn't set the field back — it revealed just how staggeringly complex human sight really is, and it set the stage for decades of deeper research. Whether you're a student exploring AI history, a developer working in machine learning, or simply curious about how technology evolved, this video offers a clear and engaging look at the milestones that shaped modern computer vision. This video was created with Ngram, an AI-powered video creation platform.