Voice Assistants and How They Work · seed 1 · A4, ink-friendly. The answer key prints on its own page for grown-ups.

Voice Assistants and How They Work

Computing · Artificial Intelligence · ages 5-7
Name ______________________   Date ____________

  1. In our spoken-answer model, what follows listening?

    • Speak the answer to the question
    • Work out the request
    • Start listening for the question

  2. In our model for answering a spoken question, what comes first?

    • Listen to the words
    • Speak a reply
    • Look for an answer

  3. A voice assistant is a real person hiding inside the speaker.

    Circle one:   True   False

  4. In our spoken-answer model, what follows finding the answer?

    • Speak the answer aloud
    • Go back to listening
    • Work out the request

  5. Which response does not match a request for today's weather?

    • It gives a weather report
    • It asks which town you mean
    • It plays a song

  6. Zoe's voice assistant says, 'I'm happy.' What does that tell us?

    • It has the same happy feeling as Zoe
    • Its software produced those words
    • A happy person is inside the speaker

  7. If a voice assistant says an answer aloud, the answer must be true.

    Circle one:   True   False

  8. Dad speaks very quietly and gets an unexpected reply. What is one possible explanation?

    • The reply proves Dad asked for music
    • The assistant felt bothered by Dad
    • It may have misheard the words
LightMySky · lightmysky.comW1-mt_u6SYiVx7FX-s1

Answer key

For grown-ups. Fold this page away before handing over the rest.

Voice Assistants and How They Work W1-mt_u6SYiVx7FX-s1

  1. Work out the request · After listening, the software works out the request before finding an answer.
  2. Listen to the words · Listening comes first in this model. It gives the software spoken information to work with.
  3. False · A voice assistant is a computer program, not a person inside the speaker.
  4. Speak the answer aloud · Speaking follows finding an answer in this model.
  5. It plays a song · Playing a song is a different task. The mismatch alone does not tell us exactly why it happened.
  6. Its software produced those words · Software can produce those words without having human feelings.
  7. False · An assistant can answer using mistaken information. Speaking does not prove correctness.
  8. It may have misheard the words · Quiet speech can be hard to pick up. Mishearing is possible, but the reply alone does not prove the cause.
Worksheet · LightMySky