OpenAI is starting to release its new advanced voice mode for ChatGPT to a select group of ChatGPT Plus subscribers. This feature was first revealed at the GPT-4 launch event in May, but it faced criticism for its similarity to Scarlett Johansson’s voice and was delayed for safety reasons.
A more capable voice mode
At OpenAIโs event, the new voice mode demonstrated notable improvements over ChatGPTโs existing voice capabilities. OpenAI staff showcased the chatbotโs ability to adjust its storytelling in response to interruptions, highlighting the modeโs flexibility and responsiveness.
Initially scheduled for an alpha release in late June, the rollout was postponed by a month to ensure the feature met OpenAI’s high standards. During this time, the company focused on enhancing the modelโs ability to identify and reject certain types of content. Taya Christianson, an OpenAI spokesperson, mentioned that over 100 external testers, known as red teamers, were involved in evaluating the modelโs robustness. This thorough testing responded to increased scrutiny of OpenAIโs safety measures, suggesting that the delay was prudent.
OpenAI has also implemented new filters to block requests to generate music or other copyrighted audio. This measure aims to address concerns about potential misuse of the technology.
Addressing the Scarlett Johansson controversy
One major criticism during the event was that the onstage voice, โSky,โ closely resembles Scarlett Johansson’s AI character in the film “Her.” Although this voice had been part of ChatGPT before the spring demo, OpenAI removed it after Johansson reportedly sent letters inquiring about its creation. Christianson assured that the new mode will exclusively use four preset voices developed with voice actors. She emphasised, โWeโve ensured that ChatGPT cannot mimic other peopleโs voices, including those of public figures, and will block any outputs that deviate from these preset voices.โ
Looking ahead to a broader release
OpenAI plans to make the advanced voice mode available to all ChatGPT Plus users by the autumn. This phased rollout aims to provide a broader audience with access to enhanced capabilities while ensuring safety and compliance with ethical guidelines.
Introducing this advanced voice mode marks a significant step forward for ChatGPT, promising more interactive and dynamic user experiences. As OpenAI continues to refine and expand this feature, it remains committed to addressing safety concerns and maintaining high ethical standards.
