Google has introduced Gemini Intelligence which can “automate boring tasks”, so users don’t have to do them themselves and can focus on other things. The function will be sent out in waves this summer. The Pixel 10 and Galaxy S26 series get the first thing.
Gemini Intelligence will eventually be released for the entire Android ecosystem, including smartwatches, smart glasses, laptops and cars. According to Google, they spent months perfecting the automation of multi-step tasks. In other words, Gemini can independently tear off tasks even though many different steps are required to complete them.
Google mentions examples such as booking gym sessions, finding and ordering student literature, or ordering food and taxis. Let’s say you have a grocery shopping list in Keep. Open the list in Keep, start Gemini and ask the assistant to build a digital shopping cart with all the items in your online grocery store. When Gemini has finished picking on the website, you only need to confirm and pay for the order to have everything delivered to your door.
The assistant only acts on behalf of the user and stops only when the task is complete, after which the user must confirm the action. Gemini will therefore not place orders left and right without your approval. We can follow the assistant’s progress through live notifications.
Autobrowse and autofill
Starting in late June, Android devices will get a browsing assistant in Chrome that should be able to help with research, summaries and comparing content on the web. A function called autobrowse can tear off mundane and boring tasks for the user, like booking different things. Nano Banana can generate images directly in Chrome. Autobrowse has prompt injection protection and will ask for confirmation before completing sensitive tasks, such as making a purchase or publishing a post on social media.
If you study and prefer visual learning, you can ask Gemini to turn the page into an “informative infographic”. Google’s classic autofill – which fills in forms automatically – is being modernized. According to Google, the finesse will become more intuitive and intelligent thanks to Gemini and a generally greater insight into the user’s Google services. The feature is optional and can be turned on and off.
Better speech-to-text through Rambler
Artificial intelligence has brought modern variants of speech-to-text and transcription. Google’s new speech-to-text is called Rambler (“svamlare” or “pladdrare” in Swedish). Users should be able to speak naturally without having to think about how they express themselves. It is possible to mix different languages in the speech.
Rambler then extracts the essence of what you said and assembles the essential parts into a concise message. Filler words are removed. According to Google, Rambler will speak in the user’s own voice – just more polished – but that remains to be seen. Finesse seems to be part of Gboard. The speech is transcribed in real time and no recordings are stored.
With Rambler, you don’t have to worry about getting your words exactly right before you start. You can speak naturally and it will take the important parts, then fit them all together into a concise message. Rambler will clearly show you when you’ve enabled it to help convert your voice to text and audio is only used to transcribe in real-time and is not stored or saved.
Create your own widgets – generative interfaces
Google ventures into so-called generative interfaces by letting users create their own widgets – something British Nothing was first to feature in smartphones. Users can create widgets with any functionality by simply describing what the widget should do.
For example, you can tell the widget creator to make a widget that suggests three high-protein recipes each week. The widget then shows new meals with associated recipes weekly directly on the home screen.
If wind speeds and precipitation are important data to you, you can create a widget that only displays that specific data on the home screen, and so on. The widget creator will be offered for smartphones, smartwatches and Googlebooks.
Availability
The advance material Swedroid received from Google does not mention anything about the availability of the automation itself in Gemini Intelligence, other than that the functions under the umbrella will be sent out gradually during the summer. However, not all the subtleties of Gemini Intelligence will be released at the same time globally. Europe will have to wait for several of the news.
For example, Gemini in Chrome for Android will only be released in the US initially. Likewise autobrowse for Android, which also requires the AI Pro or AI Ultra subscription.