Course 7, lesson 65 of 100, Ages 10+

Music, voices and video

AI that makes sounds and movies

Like I’m 5

AI can learn how songs and voices sound, then make new ones, a bit like a parrot that can sing brand-new tunes.

The big idea

AI can now generate music, clone voices and create short videos from text. These tools learned from huge collections of songs, recordings and clips.

They open up creativity, letting anyone make a soundtrack or animate a drawing. They also raise questions: whose music or voice did the AI learn from, and is it fair to copy a real person's voice without permission?

Examples

  • Music: Describe 'a calm piano tune for studying' and get a track.
  • Voices: Audiobooks read by synthetic voices.
  • Video: A drawing becomes a short animation.

How it works

  1. AI can make music, copy the style of a voice, or create short videos from a written description.
  2. It learned from lots of recordings, so it predicts what should come next, note by note or frame by frame.
  3. Because it can copy voices and faces, never use someone’s voice or face without their permission.

Check your understanding

What should you do before using someone’s voice with AI?
Options: Ask for their permission; Nothing, it’s fine; Use it secretly.
Answer: Ask for their permission. Voices and faces belong to people. Always ask first.
What should you get before copying a real person's voice with AI?
Options: Their permission; A faster computer; Nothing.
Answer: Their permission. A voice belongs to its owner, so ask first.

Remember

AI can make sounds and videos. Respect people’s voices and faces.

Talk about it

Why should we ask before using someone’s voice or photo?

Go deeper

Audio and video generators use diffusion and transformer models trained on huge media collections. Voice cloning raises consent and fraud risks, so many tools add watermarks, and laws about deepfakes and the right to your own likeness are changing quickly.