Google I/O: Gemini gets more personal, virtual fitting room and Android for glasses

Jan Johannsen

21.5.2025

Translation: machine translated

AI is set to get better and better: This summarises Google's announcements at the start of the I/O developer conference. Nevertheless, it's worth taking a closer look.

The announcements at the Google I/O opening keynote can be categorised into three areas: Innovations that will reach the most people, new and improved AI tools for a paying audience and Android XR, the new software for glasses and headsets.

Google summarises the innovations under the keywords «Personal, Proactive, Powerful». This means that the AI can memorise more things about you with so-called agents and has more context for interactions. They ensure that the AI can take on more complex tasks such as searching for and booking tickets or appointments.

Gemini sees the world around you

A year ago, «Project Astra» was still a research project. It is now so advanced that it is available as a new camera and screen share for Gemini on Android and iOS. Put simply, the AI assistant can now see what is around you or on your screen and answer questions accordingly.

AI mode for Google Search

Google is adding an AI mode to its search function. The so-called «AI mode» has its own tab, with which you can use the Gemini chatbot to start searches that can be much more complex than conventional Google searches. The same applies to the preparation of results. The function will be available in the USA from this week. It is not yet known when other countries will be added. Extensions to include finance and sports data are also in the works.

With Deep Search, the AI mode offers a function that can link hundreds of search queries and different types of information. The search takes several minutes.

Virtual try-on

Google has developed a virtual fitting room for the shopping part of its search. You upload a photo of yourself and the AI tailors the selected item of clothing to your body. This also works with other things such as a carpet in the living room.

Live translations in Google Meet

Translations by an AI have been around for some time. Google has now integrated it into Google Meet, but no longer works with subtitles or a computer voice that repeats what is being said in another language. Instead, the person speaking is synchronised - with a slight delay. For the time being, this function is only available for paying Enterprise customers.

Android XR: AI for glasses

With Android XR, Google wants to deliver the new operating system for headsets and glasses. The hype surrounding the Apple Vision Pro has already died down, which is probably why conventional glasses are the focus. Google wants to provide the software platform that others can use for their devices. Samsung, for example, has shown a headset with Project Moohan and Google has its first partners for glasses with Gentle Monster and Warby Parker.

On stage, there was a demonstration of how Gemini works on the glasses and can see its surroundings. There is also a livestream function and live translations for conversations. Further details on Android XR and the matching headsets and glasses will be released later this year.

Create videos, images and music even better

Google has further improved its tools for creating images, videos and music. «For example, Imagen 4», the tool for creating images from text, should be able to generate text and offers new image formats with square and landscape formats.

Google has further improved its tools for creating images, videos and music.

«Veo 3» now also creates matching sound for videos, including speech. The older model «Veo 2» gets control over camera movements and can remove objects. With «Flow», both models have a new tool that can be used to create entire films and no longer just individual scenes. Each shot is followed by a new prompt, whereby the AI is able to take over characters and set designs.

With Lyria 2, Google is also updating its AI tool for creating music. It is available for Youtube Shorts, Vertex AI and Music AI Sandbox.

The latest tools cost a lot of money

Not all of Google's AI tools are free to use. You have to pay for the creative tools in particular, but also for Gemini in Chrome. You need at least Google AI Pro for 21.99 euros per month. If you want to create videos with Veo 3 and Flow, you have to book Google AI Ultra for 249.99 euros per month. However, some of the latest tools are only available in the USA for the time being and I haven't found any prices for Switzerland yet.

If you want to watch the Google I/O 2025 keynote:

Jan Johannsen

Editor

Jan.Johannsen@galaxus.de

As a primary school pupil, I used to sit in a friend's living room with many of my classmates to play the Super NES. Now I get my hands on the latest technology and test it for you. In recent years at Curved, Computer Bild and Netzwelt, now at Digitec and Galaxus. more

These articles might also interest you