Skip to content
Learn Motion
ExploreHow it worksMembership
Log in
Learn Motion

How Generative AI Creates Text, Images, and Videos from a Prompt

1The Shared Idea Behind All Generative AI2How a Prompt Becomes an Instruction3Generating Text: One Token at a Time4Generating Images: From Noise to Picture5Generating Video: Adding Time6Comparing the Three Modalities and Judging Outputs
The Shared Idea Behind All Generative AI

Generation Is Choosing, Not Retrieving

3 / 4
Take the coffee shop example. The request is the same both times, yet the names differ. If the system were looking up a stored answer, that could not happen. What is happening is that several names fit the request, and the system picked one, then picked a different one on the next run. The randomness is in the choosing, not in the knowledge. That is also why you can dial the behavior: less randomness gives you steadier, more predictable results, and more randomness gives you a wider spread of ideas.
0:00 / 0:00

Generation is a sequence of choices made from a learned sense of what is plausible. Because more than one option can be plausible, and because the choice involves randomness, the output is not fixed in advance.

Why two runs differ

Ask a text tool for a name for a coffee shop. The first run returns "The Daily Grind." The second run, same prompt, returns "Morning Ember." Both fit the request. Neither was stored anywhere. The model sampled a different plausible option the second time.

If the model were retrieving stored answers, repeating a prompt would return the identical result every time. It often does not, which is one everyday sign that sampling is happening.

Previous3 / 4Next

Learn Motion

Generate a course. Learn it properly.

Operated by Wuhan Daoyin Technology Co., Ltd.

Contact: [email protected]
Privacy PolicyTerms of Service

© 2026 Learn Motion