OpenAI has begun rolling out an advanced voice feature for its ChatGPT platform, utilizing its GPT-4o model to deliver hyper-realistic audio responses. According to a TechCrunch report dated July 30, this new capability allows users to engage in real-time, natural conversations with ChatGPT, even enabling them to interrupt the AI mid-sentence, addressing a significant hurdle in creating realistic AI interactions.
OpenAI GPT-4o advanced voice features have been rolled out to some lucky ChatGPT Plus users.
It can create background music while telling stories and works well in many languages .
Some amazing example what it can do in 🧵
1 📹 via yoimnotkesku pic.twitter.com/IeoyegCk07
— AshutoshShrivastava (@ai_for_success) July 31, 2024
The Advanced Voice Mode is currently in its alpha phase and is being made available to a select group of ChatGPT Plus subscribers. OpenAI plans a broader rollout to all premium users in the fall. This phased introduction follows controversy from a demonstration in May, where the voice feature, then called “Sky,” faced backlash for its striking resemblance to actress Scarlett Johansson’s voice. Johansson, known for her role in the AI-themed film “Her,” stated she had not authorized the use of her voice and sought legal advice to protect her likeness. OpenAI denied using her voice but pulled the demo, highlighting the ongoing legal complexities surrounding AI technology and celebrity likeness rights.
To address concerns about potential misuse, OpenAI has restricted the system to four preset voices created in collaboration with paid voice actors. The company emphasized that ChatGPT is not capable of mimicking specific individuals or public figures, aiming to prevent the creation of deceptive deepfakes—a growing issue in the AI industry.
We’re starting to roll out advanced Voice Mode to a small group of ChatGPT Plus users. Advanced Voice Mode offers more natural, real-time conversations, allows you to interrupt anytime, and senses and responds to your emotions. pic.twitter.com/64O94EhhXK
— OpenAI (@OpenAI) July 30, 2024
“We tested GPT-4o’s voice capabilities with over 100 external red teamers across 45 languages,” OpenAI shared on X, formerly known as Twitter. “To protect people’s privacy, we’ve trained the model to only speak in the four preset voices and have built systems to block outputs that differ from those voices. We’ve also implemented guardrails to block requests for violent or copyrighted content.”
In addition, OpenAI has implemented filters to block requests for generating music or other copyrighted audio. This decision likely stems from recent legal actions against AI companies accused of copyright infringement. The music industry, in particular, has been vigilant in challenging AI-generated content, with lawsuits already filed against AI song-generators like Suno and Udio.
OpenAI rolling out its voice mode for ChatGPT
This new feature, exclusive to ChatGPT Plus subscribers, makes the chatbot sound like a famous actress, Scarlett Johansson. #openai #voice #AI #Scarlett Johansson pic.twitter.com/8trFIqwexB— Newsinbots (@newsinbots) July 31, 2024
Major Points:
- OpenAI has introduced an advanced voice feature for its ChatGPT platform, using the GPT-4o model to provide hyper-realistic audio responses.
- The Advanced Voice Mode is currently available to a select group of ChatGPT Plus subscribers, with plans for a wider rollout to all premium users in the fall.
- The feature faced controversy during its initial demo in May for its resemblance to Scarlett Johansson’s voice, which the actress did not authorize. OpenAI denied using her voice and removed the demo.
- To prevent misuse, OpenAI has restricted the system to four preset voices created with voice actors and implemented guardrails to block content that could imitate specific individuals or include violent or copyrighted material.
- OpenAI’s new feature comes amid increasing scrutiny over AI’s use of celebrity likeness and copyrighted content, with recent legal actions taken against other AI companies in the music industry.
Susan Guglielmo – Reprinted with permission of Whatfinger News
