OpenAI launches advanced voice mode for ChatGPT, which was criticized for sounding like Scarlett Johansson, for paid subscribers



OpenAI has announced that it will provide a new feature called 'Advanced Voice Mode' for ChatGPT Plus, a paid version of ChatGPT. Advanced Voice Mode is a feature that enables users to have natural conversations with ChatGPT by voice and to have various interactions.

OpenAI releases ChatGPT's hyper-realistic voice to some paying users | TechCrunch
https://techcrunch.com/2024/07/30/openai-releases-chatgpts-super-realistic-voice-feature/



OpenAI Debuts Advanced Voice AI for Subscribers
https://www.pymnts.com/artificial-intelligence-2/2024/openai-debuts-advanced-voice-ai-for-subscribers/

OpenAI opens limited access to ChatGPT Advanced Voice Mode | VentureBeat
https://venturebeat.com/ai/openai-opens-limited-access-to-chatgpt-advanced-voice-mode-on-mobile/

OpenAI rolls out highly anticipated advanced Voice Mode, but there's a catch | ZDNET
https://www.zdnet.com/article/openai-rolls-out-new-advanced-voice-mode-heres-how-you-can-access/

ChatGPT's Advanced Voice Mode Is Here for a Select Few
https://www.howtogeek.com/openai-launches-chatgpt-advanced-voice-mode-alpha/

OpenAI rolls out advanced Voice Mode and no, it won't sound like ScarJo
https://www.engadget.com/openai-rolls-out-advanced-voice-mode-and-no-it-wont-sound-like-scarjo-200426358.html

On July 31, 2024, OpenAI announced that it has begun offering Advanced Voice Mode to a select group of ChatGPT Plus users. Advanced Voice Mode enables more natural real-time conversations, can be interrupted at any time, and senses and responds to user emotions. The Advanced Voice Mode is expected to be rolled out to all ChatGPT Plus users by fall 2024.




The 'Advanced Voice Mode' announced this time refers to the voice conversation function of GPT-4o announced by OpenAI in May 2024. Previous ChatGPT also had a voice conversation function, but it was realized using multiple models, such as 'a model that converts voice to text,' 'a model that generates reply text based on input text,' and 'a model that converts reply text to voice,' so it was not possible to realize a natural conversation like between humans. However, GPT-4o can execute the process of 'receiving input such as voice, image, and video and then replying' with a single model, making it possible to have a very smooth conversation.

However, the new voice feature added in GPT-4o was problematic because the voice called 'Sky' was similar to the voice of Hollywood actress Scarlett Johansson. OpenAI denied that it was using Johansson's voice, but later removed the problematic voice 'Sky'. In addition, it announced that it would postpone the release of the voice feature to improve safety measures.

Scarlett Johansson expresses her opinion that she is 'shocked and angry' about GPT-4o's new voice sounding similar to her - GIGAZINE



OpenAI mentioned that it is closely monitoring the usage of the advanced voice mode provided to ChatGPT Plus users. In addition, prior to the announcement, it explained that it had 'tested the voice function of GPT-4o with more than 100 external organizations speaking 45 languages.' In addition, a select group of users will receive a notification about the advanced voice mode via the ChatGPT app, and then an email with instructions on how to use it will be sent.

In January 2024, a fake voice of President Joe Biden, using voice cloning technology from AI startup ElevenLabs, was used in the election campaign, causing a major problem. In order to avoid such a situation, OpenAI is trying to avoid controversy over deep fakes as much as possible.

AI-generated 'fake voice' phone calls of President Biden are being made to many voters - GIGAZINE



Related Posts:

in Software,   Video, Posted by logu_ii