Google just announced that Gemini will soon be able to take care of some multi-step tasks on your phone, like ordering food or hailing a car, starting with the Pixel 10, Pixel 10 Pro, and the just-announced Samsung Galaxy S26 phones. It all sounds a bit like the features Apple announced for Siri at its 2024 Worldwide Developers Conference — before Apple postponed those planned features to March 2025 and which have not yet been released.
Sameer Samat, Google’s president of Android, took the stage to demonstrate how Gemini’s new agent features would work to help pull a pizza order from his busy family group chat. Samat asks Gemini to look at the chat thread to find out what to order and then place the order using the delivery app. On the screen — in a pre-recorded video that wasn’t live — you can see Gemini finding out what everyone wants from the group chat and showing it in the window. Then the user tells Gemini via voice request to complete that order and name a specific pizzeria. Gemini then clicks on Grubhub and prepares the order, all still on the screen. When the order is ready, Gemini sends a notification so the user can review it and actually hit the submit button.
Leaving aside that this situation does not seem that complicated to do on your own in the Grubhub app (or even call a pizzeria and talk to a human about it), this is potentially a big moment for agent AI. Google recently added the ability for Gemini to automatically search for users in Chrome, and being able to do something similar directly on Android seems like the next logical step; Google clearly wants Gemini to be seen as a useful agent or productivity partner, rather than just a chatbot or a series of AI models.
Assuming Gemini’s agent features also launch “soon,” as Google promises, and that Apple doesn’t pull a rabbit out of the hat, Google will also beat Apple to one of its most impressive demonstrations of Apple Intelligence — also shown only in pre-recorded videos — from this WWDC 2024 show. One feature Apple showed off would allow Siri to understand what’s on your screen and take action on it, meaning you could ask Siri to add an address from a message thread to the contact card of the person you’re texting. Apple demonstrated how Siri will be able to perform actions for you within and across apps. The company said Siri will even be able to understand your personal context, meaning you can ask her when your mother’s flight is landing and the assistant will pull information from an email and show it to you.
Almost two years later, none of this is still available. When Apple announced that the features would be delayed, the company even pulled an ad showing off the features. And based on reports from Bloombergsome features may not arrive until iOS 27.
Of course, there are still many questions about Gemini’s new abilities. They will have to actually ship. We’ll have to try them out to see if they’re as useful and functional as advertised — Google’s calling this first run “beta,” so there may be some rough edges. And we don’t know how many developers actually let Gemini crawl their apps on behalf of users, which Rod editor-in-chief Nilay Patel likes to call DoorDash’s problem. (Google says Gemini will be able to work in “select ride-sharing and food apps.”)
But Google seems to have leapfrogged Apple in a big way, and Apple now has some work to do to catch up.