OpenAI has initiated the rollout of its Advanced Voice Mode to a select group of ChatGPT Plus users. This new feature aims to provide more natural, real-time conversations, allowing users to interrupt at any time and allowing the system to sense and respond to emotions.
ChatGPT Advanced Voice Mode
The Advanced Voice Mode is designed to enhance user interaction by offering more natural and real-time conversations. Users can interrupt the conversation at any point, making the interaction feel more fluid and dynamic. Additionally, the system is capable of sensing and responding to the user’s emotions, adding a layer of emotional intelligence to the interaction.
- Advanced Voice Mode offers natural, real-time conversations.
- Users can interrupt anytime during the conversation.
- The system senses and responds to users’ emotions.
- Gradual rollout to more Plus users over the next few weeks.
- Full availability to all Plus users expected by autumn.
- Video and screen sharing features to be launched later.
- Focus on quality and safety of voice conversations.
- Model supports millions of simultaneous voice conversations.
- Low latency and high-quality voice interactions.
- Enhanced safety measures to detect and refuse certain content.
- Four preset voices created in collaboration with voice actors.
- System blocks impersonation of individuals and public figures.
- New filters to block requests for copyrighted audio generation.
- Safety techniques applied across text and voice modalities.
- Compliance with applicable laws to prevent harmful content.
- Detailed safety report to be shared in early August.
Availability and Rollout
OpenAI plans to gradually expand access to this feature over the next few weeks, with the goal of making it available to all ChatGPT Plus users by autumn. Users participating in the alpha phase will receive notifications in the ChatGPT app and emails with instructions on how to use the new feature. While video and screen sharing are not part of the current alpha, these features are expected to be launched at a later date.
Focus on Quality and Safety
Since the preview of GPT-4o voice in May 2024, OpenAI has prioritized ensuring the quality and safety of voice conversations. The model has been reinforced to support millions of simultaneous, real-time voice conversations while maintaining low latency and high quality.
Safety Measures
In June 2024, OpenAI announced efforts to improve the model’s ability to detect and refuse certain content. Over the past few months, the model’s voice capabilities have been tested by more than 100 external red teamers, who collectively speak 45 different languages and represent 29 different geographies. Based on their feedback and internal safety tests, several measures have been implemented:
- Advanced Voice Mode uses four preset voices created in collaboration with voice actors to respect users’ privacy.
- ChatGPT cannot impersonate other people’s voices, including public figures, and will block outputs that differ from the preset voices.
- New filters have been added to recognize and block requests to generate music or other copyrighted audio.
- Safety techniques have been applied across both text and voice modalities to ensure compliance with applicable laws and prevent the generation of harmful content.
Future Plans
OpenAI plans to share a detailed report on the work done to make the Advanced Voice Mode experience safer and more enjoyable for everyone in early August. The alpha phase is focused on testing, learning, and refining the feature to deliver the best possible experience. By launching gradually, OpenAI can closely monitor usage and continuously improve the model’s capabilities and safety based on real-world feedback.
For those interested in the broader implications of this technology, other areas worth exploring include the potential for voice interaction in customer service, the integration of voice features in educational tools, and the ethical considerations surrounding AI and voice technology. These topics offer a deeper understanding of how advanced voice modes can be applied across various industries and the challenges that come with it. Here are a selection of other articles from our extensive library of content you may find of interest on the subject of ChatGPT :
- 10 ChatGPT prompt writing styles from Beginner to Master
- The world is not ready for ChatGPT-5 says OpenAI
- How to use ChatGPT 4 to fully automate your content creation
- 17+ ChatGPT advanced brainstorming prompts and concepts
- ChatGPT-5 will arrive sometime during 2024 – Sam Altman
Latest Geeky Gadgets Deals
Disclosure: Some of our articles include affiliate links. If you buy something through one of these links, Geeky Gadgets may earn an affiliate commission. Learn about our Disclosure Policy.
