Skip to main content
  1. Home
  2. Computing
  3. News

ChatGPT’s Advanced Voice Mode now has a ‘better personality’

Add as a preferred source on Google

If you find that ChatGPT’s Advanced Voice Mode is a little too keen to jump in when you’re engaged in a conversation, then you’ll be pleased to know that the latest update should bring an end to such unwanted interruptions.

OpenAI post-training researcher Manuka Stratta said in a video posted on Monday that the update gives Advanced Voice Mode a “better personality,” adding that the AI-powered tool will now “interrupt you much less … [and] because it interrupts you less, you’ll be able to have more time to gather your thoughts and not feel like you have to fill in all the gaps and silences all the time.”

Recommended Videos

Stratta added that the update also gives ChatGPT’s voice mode a more engaging and natural tone, making it more direct and concise.

OpenAI released Advanced Voice Mode for select ChatGPT Plus subscribers in May last year before expanding access to all Plus users in September. Then, last November, OpenAI rolled out Advanced Voice Mode to all Pro and free users, giving everyone the opportunity to interact with ChatGPT using voice input through the mobile app.

To use Advanced Voice Mode, simply open the app and tap on the sound-wave icon that appears on the screen beside the microphone icon. You’ll then see a circle appear on the screen, indicating that ChatGPT is listening. You’re now ready to engage in a conversation with “someone” that sounds very much like a real person but, don’t forget, is actually generative AI doing its thing. Also, to ensure that your privacy is properly protected, avoid sharing highly personal information during your chats.

Free users use an Advanced Voice Mode powered by the GPT-4o mini AI model, with usage limited to a length of time that can change daily. Those on the $20-a-month Plus tier can have much longer conversations, and use the more advanced GPT-4o model, which should produce better results in terms of elements like understanding, response quality, and reasoning. Subscribers to the $200-a-month Pro tier, meanwhile, can chat to their heart’s content, though you might want to keep an eye on just how much time you spend chatting with your AI-powered buddy.

Trevor Mogg
Contributing Editor
Not so many moons ago, Trevor moved from one tea-loving island nation that drives on the left (Britain) to another (Japan)…
Googlebook may use older Snapdragon chips to build more affordable laptops
Snapdragon X Plus and X Elite models could be in development
Googlebook

Google has already confirmed that Googlebook will support more than one chipmaker, and a new leak may have revealed the first Snapdragon models in development.

New code references uncovered by GbookHub reportedly link two Googlebook designs, codenamed Annite and Pic, to Qualcomm’s Snapdragon X Plus X1P-42-100 processor. Both are also said to be tied to a board platform called Mica.

Read more
Google’s Gemini app adds a toggle to disable AI watermarks, with some exceptions
Your AI creations no longer need a visible stamp.
nano banana

If you have ever generated an image with Gemini and gotten annoyed by that stamped-on watermark, Google finally heard you. Gemini is rolling out a new setting called Media Watermarks that lets you toggle visible watermarks on or off across everything you create, including images made with Nano Banana, videos made with Omni, and music made with Lyria. Until now, that visible watermark was applied automatically with no way to turn it off.

https://twitter.com/geminiapp/status/2088277477672255830?s=46

Read more
U.S. courts will now make government use of spyware tools public
U.S. courts will soon start putting numbers on government spyware use
landfall-spyware-hacker

U.S. courts are set to begin publicly reporting how often federal judges authorize the government to use spyware and hacking tools to intercept real-time communications, giving the public its first standardized view of how frequently these surveillance techniques are being used.

The change will begin with data collected for the 2028 Wiretap Report, which will be published in 2029. The Administrative Office of the U.S. Courts told TechCrunch that the report will introduce a new “spyware/hacking” category covering wiretaps conducted using hacking tools and spyware, known to federal authorities as network investigative techniques, or NITs.

Read more