This new tech puts AI in touch with your emotions

Hume AI's new empathic voice interface, EVI 2, brings emotional intelligence to AI models.


News Desk September 12, 2024
Courtesy: Reuters

Listen to article

Hume AI, a New York-based startup, has unveiled an empathic voice interface called EVI 2, a voice-to-voice foundation model designed to transform how we interact with artificial intelligence (AI).

This latest development represents a significant leap forward in making AI more emotionally expressive and attuned to user emotions.

By integrating with major AI platforms like Anthropic, Google, Meta, Mistral, and OpenAI, EVI 2 sets the stage for emotionally aware AI interactions.

At its core, EVI 2 enables AI to communicate with a remarkably human-like quality.

It can not only understand the tone of a user’s voice but also respond in kind, offering a more natural and emotionally engaging experience.

What makes this model stand out is its ability to generate a wide range of emotions, accents, and even speaking styles, adapting to different personalities and preferences.

EVI 2’s ability to converse fluently with users at subsecond response times is one of its key highlights. Its fast response rate helps make interactions smoother and more lifelike. Users can even make unique requests, such as adjusting the AI’s speaking speed or asking it to switch to a more entertaining style, like rapping.

In this way, the model's versatility enables it to take on a variety of personalities, further enriching conversations.

What sets EVI 2 apart is its deep training in emotional intelligence.

This model is not just designed to answer questions or execute commands—it’s optimized to anticipate and adapt to users' emotional cues. By recognizing shifts in tone and mood, EVI 2 responds in a way that feels both appropriate and empathetic.

This approach allows AI to foster a deeper connection with users, making interactions more enjoyable and engaging.

One of EVI 2’s core strengths is its ability to align with each user’s preferences.

Through a series of subtle, rapid decisions made during every interaction, EVI 2 creates a personality that feels tailored to the individual. Over time, this model is capable of evolving, learning from interactions to better understand and cater to a user’s emotional needs.

Hume AI's mission is clear: to optimize AI for human well-being. With EVI 2, the focus is on creating positive and satisfying experiences for users, making the AI feel more personable and relatable. According to the company, this emotional alignment could be a key factor in increasing overall happiness and satisfaction in AI-assisted tasks.

In addition to its emotional intelligence, EVI 2 possesses impressive multilingual capabilities. This AI model can communicate in various languages, making it accessible to users worldwide.

Moreover, it can seamlessly switch between languages, accents, and speaking styles based on the user's preferences or cultural background.

EVI 2 also offers multimodal functionality, enabling it to integrate voice recognition and response across different applications. Whether it’s for customer service, personal assistants, or gaming, EVI 2 can be customized to fit a wide range of use cases.

Currently, EVI 2 is available in beta for public use through the Hume AI app. Developers can also access its API to integrate the model into their own applications. This accessibility allows businesses and individuals alike to experience the future of AI communication firsthand.

By focusing on emotional intelligence and adaptable personalities, EVI 2 opens new possibilities for AI interactions. As AI continues to evolve, the ability to understand and respond to emotions may be a crucial factor in its widespread adoption.

Hume AI believes that emotionally intelligent AI will play a key role in enhancing human-AI interactions. The company’s ongoing research focuses on fine-tuning the model to optimize for happiness and user satisfaction.

COMMENTS

Replying to X

Comments are moderated and generally will be posted if they are on-topic and not abusive.

For more information, please see our Comments FAQ