OpenAI is making headlines once again with the rollout of ChatGPT's Advanced Voice Mode. Starting today, a small number of ChatGPT Plus subscribers will get their hands—or rather, their ears—on this cutting-edge feature. This new mode, powered by the GPT-4o model, promises to deliver hyperrealistic audio responses that are set to redefine our interactions with AI.
A Voice That Sounds Real
When OpenAI first showcased GPT-4o’s voice capabilities back in May, it was nothing short of astonishing. The demo highlighted the model's ability to produce quick, natural-sounding responses. However, one voice, named “Sky,” caught everyone's attention for its uncanny resemblance to actress Scarlett Johansson. Johansson, known for her role as an AI in the movie "Her," publicly stated that she had declined OpenAI's requests to use her voice. The controversy led to legal action and a subsequent decision by OpenAI to remove the voice from its lineup. The rollout of Advanced Voice Mode was delayed by a month to ensure the necessary safety measures were in place.
Real-Time Conversations with Emotional Nuance
GPT-4o stands out for its multimodal capabilities, seamlessly integrating voice, text, and vision. Unlike previous iterations that relied on separate models to handle voice and text conversion, GPT-4o processes everything internally, resulting in significantly lower latency. This allows for more natural, real-time conversations. Additionally, the model can detect emotional tones in users' voices, such as sadness, excitement, or even singing, and respond accordingly.
A Cautious and Controlled Release
The initial rollout of Advanced Voice Mode is intentionally limited to closely monitor usage and address any issues that may arise. Selected alpha users will receive notifications in the ChatGPT app, followed by an email with detailed instructions. This careful approach underscores OpenAI’s commitment to safety and quality.
To prevent misuse, the new mode will feature only four preset voices—Juniper, Breeze, Cove, and Ember—created in collaboration with professional voice actors. OpenAI has also implemented filters to block requests for generating copyrighted audio, aiming to avoid the legal challenges faced by other AI companies.
Future Prospects
While the current alpha release focuses on voice capabilities, OpenAI has bigger plans. Features like video and screen sharing, which were teased during the Spring Update, are set to be introduced at a later date. These additions will further enhance ChatGPT’s utility, making it an even more versatile tool for users.