users still had access to it


While OpenAI says GPT-4o’s Advanced Voice feature won’t arrive for several months, some users say they’ll have access to it anyway. Enough to discover concretely what this new version of the voice assistant looks like.

Source: Unsplash

Last May, OpenAI presented Advanced Voice mode, a ChatGPT function that allows you to chat transparently with the chatbot, like a voice assistant. An announcement which was accompanied by GPT-4o, a multimodal LLM more powerful than GPT-4. OpenAI indicated that this was not going to happen very soon: we will have to wait. However, several users still have access to it, as reported Tom’s Guide.

A ChatGPT feature that no one should have had access to

One of the user examples is a certain RozziTheCreator, who, on Reddit, shared the recording of a conversation with ChatGPT. We hear a new voice that is currently not available and that we have never heard before. A voice that tells a horror story, adding disturbing sound effects. As this Internet user recounted, there was only access for a few minutes, and the tool encountered a certain number of technical problems.

This content is blocked because you have not accepted cookies and other trackers. This content is provided by Reddit.
To be able to view it, you must accept the use made by Reddit with your data which may be used for the following purposes: allowing you to view and share content with social media, promoting the development and improvement of products from Humanoid and its partners, display personalized advertisements to you in relation to your profile and activity, define a personalized advertising profile, measure the performance of advertisements and content on this site and measure the audience of this site (find out more more)

By clicking on “I accept all”, you consent to the aforementioned purposes for all cookies and other trackers placed by Humanoid and its partners.

You may withdraw your consent at any time. For more information, please read our Cookie Policy.

Manage my choices

Asked by Tom’s GuideOpenAI has indicated that it was indeed a bug, since fixed, that caused some users to have access to the voice version of GTP-4o. A rather good mistake for us, less so for OpenAI: this is the first time that we discover this Advanced Voice mode not directly controlled by OpenAI. Until now, we have only been treated to demonstrations, under the control of its developer. What RozziTheCreator’s video shows is an AI that speaks in a fairly natural way, and that adds sound effects: this had never been seen before.

Advanced Voice mode with GPT-4o lags behind

It was in a publication on X that OpenAI expressed its progress with the voice mode of ChatGPT under GPT-4o. The company recalled that it had planned to start deploying this function in a first version to some users of ChatGPT Plus. However, everything is not ready: “We are improving the model’s ability to detect and reject certain content. We are also working on improving the user experience and preparing our infrastructure“. This is why the launch is delayed by a month.

A demonstration of ChatGPT Voice // Source: OpenAI

There is therefore a delay, which means that all ChatGPT Plus users will have access to Advanced Voice in the fall: “Exact delivery times depend on meeting our high safety and reliability standards.»



Source link -102