Gemini Live might be the conversational AI that sci-fi promised us
Google's expanded virtual assistant is getting chatty
The tech world has been obsessed with AI smartphones for a while, but it still feels like we’re a long way from “Joaquin Phoenix falls in love with a gadget” levels of lifelike. Google’s new, more conversational way to interact with its virtual assistant, Gemini Live, might be the next closest thing.
Instead of saying a wake word every time you want to chat with your phone, Gemini Live is bit more like a phone call you can pause at any time. You can continue a convo with the Gemini app in the background, or when your phone is locked. It’ll work hands-free through your connected wireless earphones, too.
There’ll be ten different voices to pick from, which Google says will help make speaking to Gemini a little more natural. Google reckons this back-and-forth AI banter will be brilliant for brainstorming ideas, practicing important conversations and pulling data from a bunch of different apps.
The latter uses Extensions, which will integrate Gemini into Google’s biggest apps. Eventually it’ll be able to check for calendar clashes, set reminders and add to your Google Keep task lists. You’ll also be able to ask it to create YouTube Music playlists, and control device actions like Bluetooth and volume.
Other AI upgrades include being able to recognise text from your photo library, so you can extract shopping lists from recipes, or check if your diary is free from a gig flyer. Gemini will also be able to create generative images from your voice prompts, then let you drag and drop them into Google Messages or Gmail.
Launched today alongside the Pixel 9 series phones, Pixel Watch 3 and Pixel Buds Pro 2, Gemini Live is heading to English speakers on Android first. More languages and iOS support is due over the coming weeks. You’ll need a Gemini Advanced subscription – part of the $20/£19 Google One AI Premium paid plan – to get it.
Tom Morgan-Freelander
Deputy Editor 
