How do generative AI algorithms work?

How does the information a generative model knows about the world affect the types of images it can create?

Objectives:

• Explain how AI algorithms are able to generate content that is based on, but not identical to, the content they were trained on

• Describe the effect that training data has on generative models and discuss how biases in training data can affect the effectiveness of the model

• Evaluate how realistic modern generative AI models can be and where they still have room for improvement

Source: CRAFT - Stanford University