Generative Models
AI systems that create new data—like images, text, or video—by learning patterns from training data.
In Plain English
Generative Models are trained to understand the underlying patterns in data and then produce new, original examples that resemble what they learned from. Unlike systems that classify or predict existing categories, generative models actually create something novel. They work by learning the structure and rules buried in their training material—how pixels relate in photos, how words flow in sentences, how musical notes build melodies—and then using that knowledge to generate fresh outputs. They power tools like image generators, text writers, and code assistants.
💡Real-World Example
DALL-E and Midjourney are generative models. When you describe a scene in words—"a cozy coffee shop in autumn"—the model creates a brand new image it has never seen before, combining patterns it learned from thousands of real photos and the relationship between descriptions and visual elements.
Related Terms
What did you think of our explanation?