BInvestigatorAudio and speech· about 20 min· fundamentals that rarely change· verified 2026-09-20· EN
How does the voice understand me?
Be able to describe the steps from speech to text to answer at a simple level.
Prerequisites
Everyday explanation
You say «What's the time?» to a voice assistant. What happens?
- The microphone turns the sound wave into numbers.
- Speech recognition guesses which words were said: «what's the time».
- Language understanding works out what you want: to know the time.
- The answer is fetched: 14:32.
- Speech synthesis turns the text back into sound: «It is fourteen thirty-two.»
Five steps, five places where it can go wrong.
Interactive
Fool the assistant: try saying «What's the time?» with chewing gum in your mouth, with music in the background, in a strong accent. Which step went wrong — did it hear the wrong words (step 2) or understand them wrongly (step 3)?
Say «Set a reminder to set the washing going» — «set» appears twice in different roles; the assistant has to understand the sentence, not just the words.
Think about this too: the assistant listens for its name the whole time. What do you make of that?
Mastery means
- Describes the steps speech → text → answer → speech
- Gives an example of where in the chain it can go wrong
Sign in to do the exercises and build your mastery up.
Sources
- Swedish Wikipedia — Speech recognition (CC BY-SA 4.0) — CC BY-SA 4.0
- Skolverket — About AI in school (in Swedish) — Skolverket's open terms