Gemini’s biggest innovation is all about controlling your phone

It is, again, the season of Gemini. Google is announcing new Gemini features at its pre-I/O Android event, many of which are aimed at helping you use your phone. You’ll find Gemini in many places, like Chrome on Android, in your auto-loader view, and all in your apps – if you want.

Google also has a new name to remember, because I can’t help myself: Gemini Intelligence. “It brings the best of Gemini to our most advanced Android devices,” according to Google’s head of Android, Ben Greenwood. Google is combining some of the existing and new Gemini products under this name, and it seems to be reserved for high-end Android phones such as the Galaxy S26 series. Write that on your I/O bingo card.

Task automation is clearly one of the “best of Gemini” features. I am they are on the latest Pixel and Samsung Galaxy phonesand enables Gemini to run other programs on your behalf. It has been limited to food and nutrition programs until now. That’s changing “soon,” says Google, when the operating system will open up to more apps.

It will also increase multimodality; in the past, Gemini could use words or phrases to inform their actions. Now you can throw a picture or a mixed picture, which seems like something you should do from the beginning. You will be able to give Gemini a picture of the shopping list in your notes and it will add the items to your cart. You know, if you have an Android phone that supports Gemini Intelligence.

Another feature under the Gemini Intelligence umbrella that is new is Create My Widget. Google’s blog post calls it the first phase of “design UI,” and it lets you describe the functionality you want in natural language and let AI create a custom widget. Google’s examples include weather for a cyclist who wants to see wind speed and rain in real-time, and a dashboard for cooking ideas, such as “three great breakfast recipes every week.” These widgets will also make their way to Wear OS, so they’ll be available on the watch, too.

It’s a simple idea on the face of it, but if you think of widgets as little apps that you can dance on your phone’s screen, it’s a lot more interesting. Maybe this is actually a small step to the form that just it creates itself on the fly. Or maybe it’s too complicated to install a humble widget. In any case, I’ll be interested if we hear more about the main I/O topic about “design UI.”

Google is also bringing the Gemini features available on the Chrome desktop to its Android app. This means you’ll see a Gemini button in Chrome where you can share content and ask Gemini questions right inside the browser. If you’re a subscriber to Google’s AI Pro or Ultra plans, you’ll also get temporary browsing to help you complete your tasks, like booking an appointment. This will start rolling out at the end of June.

Gemini will also appear – if desired – by auto-loading on Android. You will be able to choose to connect Gemini to help you fill out forms. This means that Gemini can use it its connection to Personal Intelligence to things like Google Photos and Gmail to see more relevant information. In theory, this could mean things like pulling your license number from Photos. Helpful? Dangerous? Some combination of all these? It’s Gemini season, baby. The Gemini Intelligence feature “will appear to be ready this year,” says Greenwood, whose Galaxy and Pixel phones are the first to get the update this summer.