During Google I/O 2024, the tech giant unveiled Gemini Live, a unique take on mobile AI capable of holding “natural conversations.” You can think of it like Siri, as you’ll be able to engage in a two-way dialogue with Gemini and have the AI respond concisely.
Since the initial announcement, we hadn't heard much about the technology until recently, when strings of code referencing Gemini Live were found in a beta version of the Google app. The code mentions a “background mode” that allows users to continue conversations “while using other apps or while [their] “The screen is locked.” As 9To5Google points out, this effectively turns the Gemini Live experience into a sort of phone call.
AI will still accept voice input and respond while you do other things. A user could, for example, open Chrome on their Android phone, search for something in the search engine, and seamlessly relay that information back to Gemini. Again, the goal is to make talking to AI second nature.
9To5Google says interactions can be ended by “looking for a persistent notification” or saying “Stop” to Gemini Live out loud.
Possible characteristics
However, there was more to the beta than just background mode.
Google is apparently working on a way to give users quick access to Gemini when their smartphone is locked. If you look at the screenshot in the report, you'll be able to use four extensions for its own services: Google Maps, Google Flights, Google Hotels, and YouTube.
It's not known exactly what the extensions will be able to do, as details are scarce. But judging by what was shown, users could ask Gemini for directions, play music on YouTube or book a flight.
Those four extensions will be enabled by default, and Google apparently has plans to add support for more extensions; however, there is a small catch. According to the screenshot, “Extensions that don’t require user credentials or access private data will be enabled by default.” Everything else would presumably require manual activation.
Additionally, users can toggle three additional Gemini features on and off from the lock screen, expanding the level of AI control. Messages allows texting friends and family, while Workspace can locate and summarize emails or documents. Then there’s Home Automation, which is arguably the most interesting, as it allows homeowners to manage smart home devices from the lock screen.
Surpassing Apple
There's no word on when Gemini Live will be released. The May announcement said the software would be coming to Gemini Advanced subscribers in the coming months. So far, Google hasn't said anything. We might see it before the end of the summer or later in the fall.
When it launches, background mode and lock screen interactivity could allow the update to stand out against the likes of Apple Intelligence. Apple's technology is also very capable, offering all sorts of tools for organization, content creation, and more. Judging from what we've seen, though, it requires constant interaction. Gemini Live is less strict, which could make it easier to use.
As always, these leaks need to be taken with a grain of salt. Things can always change at the last minute.
If you're interested in knowing which is the best AI chatbot out there, check out our Versus article where we compare the big three: Meta AI, ChatGPT, and Google Gemini.