Google is testing voice-customization features for its Gemini AI assistant that allow users to adjust speaking speed, formality, warmth, and energy [1, 2].
These adjustments represent a shift toward more personalized human-computer interaction. By moving away from static presets, Google aims to make AI assistants feel less robotic and more aligned with individual user preferences.
The proposed system would likely provide users with controls to fine-tune how the assistant sounds during interactions [1]. This includes the ability to modify the energy level of the voice, which could range from a calm, subdued tone to a more enthusiastic delivery. Users could also adjust the warmth of the AI, potentially making the assistant sound more friendly or more professional depending on the context of the task [2].
Formality is another key pillar of the testing. A user might prefer a highly formal tone for business-related queries but a more casual, relaxed style for home automation or entertainment requests [1]. The addition of speaking speed controls allows users to accelerate the delivery of information or slow it down for better comprehension [2].
Google has not yet announced a wide release date for these specific customization sliders. The company is currently evaluating how these changes affect the overall user experience and the naturalness of the AI's speech patterns [1, 2].
This move follows a broader industry trend where AI developers are attempting to bridge the gap between synthetic voices and human emotion. By giving users direct control over the "personality" of the voice, Google is attempting to reduce the friction often associated with voice-activated technology [2].
“Google is testing voice-customization features for its Gemini AI assistant”
The shift toward granular voice customization indicates that AI developers are moving beyond simple text-to-speech accuracy and into the realm of emotional intelligence. By allowing users to define the 'personality' of the AI, Google is acknowledging that a one-size-fits-all voice is insufficient for a global user base with diverse cultural and situational needs.
