release

How the voices for ChatGPT were chosen

For AI workflow builders, the voice selection process underscores the need to prioritize user experience and accessibility in voice interfaces, which can significantly impact adoption and satisfaction in deployed AI applications.

OpenAI Blog·May 19, 2024·1 min readrelease

releaseHow the voices for ChatGPT were chosen

openai.com

What happened

OpenAI has disclosed the process behind selecting the five voices used in ChatGPT's voice mode. According to the OpenAI Blog, the company collaborated with professional casting and directing experts to evaluate over 400 submissions before finalizing the five voices. This selection process aimed to ensure high-quality, natural-sounding speech interactions. The choice of voices is a key component of ChatGPT's multimodal capabilities, which allow users to interact via speech rather than text alone. For developers and solopreneurs building AI workflows, understanding this curation process highlights the importance of user experience design in voice-enabled AI systems. The voices were chosen to be broadly appealing and unintimidating, reflecting OpenAI's focus on accessibility and usability. This editorial does not endorse any specific voice but notes that the selection process was rigorous and based on industry best practices.

Key takeaways

OpenAI worked with industry-leading casting and directing professionals to evaluate over 400 voice submissions.
The company selected five voices for ChatGPT's voice mode after a thorough curation process.
The selection aimed to make voice interactions natural, high-quality, and accessible to a wide audience.
The voices are a part of ChatGPT's multimodal features, enabling speech-based interactions.