Voice

Voice interfaces take away the thing most UX relies on: a screen to scan. A smart speaker, an in-car assistant, a phone dictation mode, or a spoken chatbot delivers its output as a single linear stream that is gone the moment it ends. People cannot skim ahead, glance back, or compare options side by side, so everything has to fit in working memory. Input is uncertain too: every request passes through speech recognition, and the system has to act on its best guess of what was said.

What separates good voice interfaces

They pick tasks that suit speech — quick questions, timers, hands-busy commands, short dictation — and leave long comparison to a screen. Prompts are short enough to hold in your head. Anything with consequences, such as sending a message, buying something, or deleting data, is read back and confirmed before it happens. Good voice products tell people what they can say instead of waiting for them to guess, and they signal plainly when they are listening, thinking, or done. The words carry the whole design, which makes UX writing and clear feedback and system status the core skills here, and a visual or text fallback keeps the product usable for people who cannot or would rather not speak — basic accessibility.

Where voice UX usually breaks

Recognition errors with no recovery path, where a misheard name sends the conversation down the wrong branch and the only fix is starting over. Spoken menus that list six options when nobody can remember more than a few. Features that exist but are never mentioned, so people stick to the three commands they already know. An unclear listening state that leaves users talking to a device that stopped listening. Context lost between turns, forcing people to repeat what they just said. And privacy worries around wake words, recordings, and replies spoken aloud in shared rooms.

This hub collects TYPENORM's voice articles. For examples in the catalog, compare how Pi speaks its replies aloud in a chosen voice, voice input in Gemini and Google Search, and the hold-to-record voice notes in WhatsApp.