ChatGPT 'Voice Mode' that lets you talk to AI like human lands next week
Technology
But GPT-4o is much smarter and, importantly, far quicker
(Web Desk) – ChatGPT’s long-awaited upgraded Voice Mode powered by the company's smartest AI system is nearly here.
Chatbot creator OpenAI said that the new voice features – which allow for spoken humanlike conversations – will roll out next week.'
It's all powered by GPT-4o, a large language model that was unveiled back in May.
As part of the demo, OpenAI showed off stunning demos of the new Voice Mode in action.
In one particularly impressive AI feat, the chatbot was able to translate in real-time for two people speaking different languages – letting them hold a conversation.
OpenAI's GPT-4o is available right now for text-based conversations, and the upgraded Voice Mode is coming next week.
That's according to OpenAI chief Sam Altman, who shared that the feature would only be available to ChatGPT Plus subscribers.
That's a paid-for version of ChatGPT that costs $20 a month.
It'll give you access to the "alpha" version of the feature, which means it's still in testing.
It's important to note that ChatGPT already has a voice mode that's powered by the older GPT-4 model.
But GPT-4o is much smarter and, importantly, far quicker.
There's a small delay when voice-chatting with ChatGPT using GPT-4.
That will almost disappear entirely when you're using the new GPT-4o version.
"It can respond to audio inputs in as little as 232 milliseconds, with an average of 320 milliseconds," explained OpenAI.
"Which is similar to human response time in a conversation."
That's compared to Voice Mode with GPT-4, which had average delays of 5.4 seconds.
You'll be able to ask the GPT-4o to talk in different tones of voice or change its pace.
And it can even translate foreign languages in real time, acting as a translator between two people.
This means two people who speak totally different languages can have a conversation using the app's Voice Mode.
