05.15.2023 | Warm Up | Review 4.5

Which of the following tasks are examples of generation? (Select all correct answers.)

• Creating new songs from existing samples of songs and music

• Creating new paintings from existing samples of images of paintings

• Deciding if a piece of mail should go to a specific address or not

• Creating new poems from existing samples of poems

Which of the following tasks are examples of classification? (Select all correct answers.)

• Deciding if an image of a dog is a poodle or not.

• Deciding whether something is recyclable paper or not.

• Creating an image of a Persian cat from other images of Persian cats.

• Sorting ripe avocados from unripe avocados.

Images and photographs cannot be considered true art.

• True

• False

Which of the following might happen if a GANs was trying to create a fake image of 
 Mount Everest? (Select all that are true)

• The generator algorithm creates new images of mountains to show the discriminator

• The discriminator algorithm judges each mountain image the generator algorithm presents and decides if it is a fake (or not).

• A discriminator algorithm gives feedback to the generator algorithm on why it thought it was a fake (so the generator can improve).

• The GANS would just be able to read our minds, see the image of Mount Everest in our imaginations, and create it!

If a generative AI system was never shown images of cats, it might have trouble generating new images of cats.

• True

• False

In a GAN there are two algorithms at work – one is a “generator” (that creates new content to show) and the other is a “discriminator” (that decides if the content it sees is fake or real).

• True

• False

GAN stands for “Generative Adversarial Network”

• True

• False

Which of the following types of content can be created by generative AI systems? 
 (Select all correct answers.)

• Text

• Images

• Video

• Audio (sounds)

Generative AI systems cannot create new images.

• True

• False

Generative AI systems can generate new content based on data provided to it.

• True

• False