In a groundbreaking development, OpenAI has announced the release of a voice assistant feature that mirrors the emotional intelligence portrayed in the film Her. This new technology is capable of reading facial expressions and translating spoken language in real time, providing an interactive experience that closely resembles human interaction.
Enhanced interactivity and responsiveness
On May 13, during a live demonstration, OpenAI’s team, including CTO Mira Murati, showcased the innovative features of this assistant. The presentation highlighted the assistant’s ability to engage creatively, such as by narrating a bedtime story with varied emotional tones, including switching to a robotic voice and concluding with a song. Furthermore, the assistant demonstrated its capability to interact with the environment by responding to live camera feeds and effectively acting as a real-time translator without requiring repeated prompts.
This level of interaction is notably similar to the character voiced by Scarlett Johansson in Her, suggesting an advanced AI that can form complex interactions with humans. OpenAI CEO Sam Altman, after the demonstration, made a brief yet telling post on X, simply stating “her,” alluding to the inspiration behind this innovation. Her explores the depth of human-AI relationships, and this new assistant aims to encapsulate that emotional connection.
Clarity on voice comparisons
During a briefing with The Verge, Murati addressed comparisons to Johansson’s voice, emphasising that while the assistant’s voice is highly expressive and personable, it is not specifically modelled on Johansson. “This might be why it feels so distinct from the more mechanical tones we’re used to with ChatGPT,” Murati explained.
This advancement significantly improves upon the existing capabilities of ChatGPT’s voice mode, which does not support interruptions or environmental interactions. The upcoming “alpha” release, set for the coming weeks, will initially be available to ChatGPT Plus subscribers, with plans for a broader rollout later.
Potential integration with Apple
The announcement coincides with Bloomberg reports about a potential partnership between OpenAI and Apple to integrate ChatGPT into the iPhone ecosystem. While Murati did not confirm these discussions, the integration could potentially solve Apple’s Siri’s long-standing reliability issues, providing a more intuitive and responsive assistant.
Sam Altman expressed his enthusiasm about the project in a blog post, stating, “The new voice and video mode is the best computer interface Iโve ever used. It feels like AI from the movies, and itโs still surprising that itโs real.” He emphasised that achieving human-like responsiveness and expressiveness in AI is a transformative milestone.
OpenAI’s latest release is set to redefine the boundaries of human-AI interaction, promising a future where digital assistants are not just tools but partners capable of understanding and responding to human emotions.
