OpenAI reveals how ChatGPT voices were chosen

OpenAI’s ChatGPT introduced the Voice Mode nearly a few months ago. ChatGPT had incorporated five distinct voices into its voice technology by September 2023- Breeze, Cove, Ember, Juniper, and Sky. Each voice received positive feedback from the users and OpenAI recently shared how they selected these for the large language model. The journey of five months behind this development included a close collaboration with professional voice actors, talent agencies, casting directors, and industry advisors.

Let’s dive into the details.

OpenAI Releases Its First AI-Made Music Video

About ChatGPT’s Voice Capabilities

OpenAI has added voice features to ChatGPT, giving users a different and interactive method to communicate with the AI. After its release, people’s reactions have been very positive. They especially like how every voice has its distinct personality.

To ensure proper casting of the voices, OpenAI actively collaborated with voice actors, talent agencies, and casting directors. Each voice actor involved got compensation that went beyond the rates found at the top of the market. It was significant to them that their AI voices should not be imitating famous personalities. OpenAI makes known that Sky’s voice isn’t a copy of Scarlett Johansson but rather the natural voice of a professional actress.

OpenAI to Soon Launch its Search Engine to Rival Google

How ChatGPT voices were chosen?

At the beginning of 2023, OpenAI made a team with famous casting directors and producers. They set some qualities that ChatGPT’s voices should have. The goal of this teamwork was to find voices that are timeless and friendly, also sparking trust and interest.

In May 2023, a call for casting was put out. There were more than 400 voice and screen actor submissions. These actors recorded scripts that covered various topics like mindfulness, travel planning, or simple day-to-day conversations as ChatGPT responses.

Key characteristics they sought in the voices included:

• Diverse Backgrounds: Actors from various cultural backgrounds or those who could speak multiple languages.
• Timeless Quality: Voices that feel timeless and can appeal to a broad audience.
• Approachability and Trust: Voices that are engaging, confidence-inspiring, and charismatic.
• Rich Tone and Warmth: Voices with a warm, engaging, and rich tone that sounds natural and pleasant to the ears.

The recording of auditions was reviewed by the casting agency and directors, who assessed the submissions. After a thorough evaluation, a total of 14 actors were selected from the talent pool. They went through further review to decide if they fit into final roles.

The chosen voice actors were given a detailed explanation by the company about their plans for human-AI voice interactions. Conversations talked over the abilities of technology, what it can’t do, and how to keep things ethical.

Haiper AI Text to Video Generation: How to Use? Is it better than OpenAI Sora?

Production and Roll Out

In June and July, the actors selected for this work were asked to come to San Francisco for recording. These meetings took place with the OpenAI product and research teams, giving the chance to meet in person for immediate feedback and changes needed. The main goal was to document natural subtleties and sound qualities that would give each voice its own distinctiveness and appeal.

September 25, 2023, was the day when ChatGPT officially opened its Voice Mode to Breeze, Cove, Ember, Juniper, and Sky. This started a new era in user interactions with more natural and lively conversations that elevated the overall experience.

Meet the Young Indian Behind OpenAI’s GPT-4o Innovation

Future of Voice Technology in ChatGPT

For the future, OpenAI is preparing a Voice Mode for the recently introduced GPT-4o. It will be available in alpha at first to ChatGPT Plus users. This advanced mode promises more natural conversation, better handling of interruptions and group talks, noise filtering, and tone-matching capabilities. The company has faith in these improvements, believing they will make voice interactions smoother and more natural.

Additionally, the AI research and development company plans to enhance the range of voices on ChatGPT. They are working towards providing various options that meet different user tastes and likes. This will contribute to making ChatGPT more flexible and interesting.

Who Is Google Veteran Shivakumar Venkataraman Hired By OpenAI?