At the Google I/O 2025 event, Google unveiled an innovative feature for Google Meet: real-time speech translation. This advancement, powered by a sophisticated language audio model developed by Google DeepMind, facilitates seamless conversations between individuals speaking different languages.
The speech translation feature works by converting spoken words into the listener’s chosen language instantly, while also maintaining the speaker’s voice, tone, and emotional expression, ensuring a more authentic communication experience. This functionality opens up numerous possibilities; it could be beneficial for English-speaking grandchildren conversing with their Spanish-speaking grandparents or for businesses with a global workforce, enabling colleagues in different regions to engage effortlessly in real time.
Google emphasizes the remarkably low latency of this translation system, allowing for multiple participants to interact simultaneously, a capability not previously available. When one person speaks, their original voice is still faintly audible under the translated speech, adding a layer of personal touch to the interaction.
The rollout for this feature will begin on Tuesday, initially available to consumer AI subscribers in a beta version, starting with English and Spanish. More languages, including Italian, German, and Portuguese, are scheduled to be introduced soon after.
Additionally, Google is planning to enhance this translation capability for business users, with early testing set for Workspace customers later this year. This development marks a significant step forward in bridging communication gaps in our increasingly interconnected world.
Fanpage: TechArena.au
Watch more about AI – Artificial Intelligence
