Google has launched Gemini 3.5 Live Translate, an artificial intelligence that translates voice in real time across more than 70 languages. Its novelty is that it not only converts words: it preserves the speaker's intonation, rhythm, and emotion. The tool promises to break down language barriers in video calls, meetings, or travel.
How the AI handles tone and context 🎙️
The system uses an audio processing model that analyzes pauses, accents, and volume variations. Then, a synthetic voice generator reproduces the translation with patterns similar to the original ones. Google claims latency is under 2 seconds. The key lies in its neural network trained on thousands of hours of multilingual dialogues, capable of distinguishing between a sarcastic question and a direct command.
Now you'll be able to sound just as irritating in 70 languages 😤
Finally, you can argue with a German call center agent without losing your characteristic complaining tone. Or order a coffee in Japanese with the same lack of enthusiasm as in Spanish. The AI even preserves your dramatic pause before answering. The downside: if you're a pushy salesman, your cloying enthusiasm will also be faithfully translated. Get ready to be rejected in all languages with the same politeness.