Google is increasingly expanding the integration of its generative AI chatbot Gemini to various products, from Searche to Google Maps and many other apps. Gemini will also soon be available on smartwatches running WearOS and in cars via Android Auto.
Consequently, this week’s I/O developer conference is largely about the Gemini AI platform.
The latest development for AI companies are so-called agentic applications, where users can engage an AI app to gather information and perform tasks over an extended period of time. Google is working on an application under the name Project Mariner, which adds functionality to the Gemini app.
CEO Sundar Pichai (pictured) showed a demonstration where he asked Gemini to search for an apartment in Austin, Texas, for three people, with criteria such as a monthly budget and amenities such as a washer/dryer in the building. Gemini then searched websites local property websites to find suitable apartments and continued to monitor new listings.
Pichai also showed a new feature in Google Meet: live voice translation. In its demonstration, a user spoke to a vacation rental agent in South America, a convresation translated from English to Spanish (and vice versa) in real time via audio.
Translation between English and Spanish is now available to subscribers, with more languages to follow in the coming weeks.
The company tonight also unveiled its newest product, Google Beam, which Pichai described as an AI-first video communications platform.
Experimental chats
Building on the earlier Project Starline, Beam uses multiple cameras and sensors to turn 2D video streams into lifelike 3D images, making conversations feel more natural and immersive.
The system uses multiple cameras to capture different angles and converts these 2D video streams into 3D images. This allows users to make eye contact, perceive gestures and experience a sense of physical presence during video calls.
Although Google Beam is currently still in the experimental phase, it is already being tested by companies such as Salesforce, T-Mobile and WeWork. The technology lends itself to sectors such as healthcare and customer service, among others, where personal interaction is crucial.
Google also introduced Flow, a new AI tool for filmmakers that uses the advanced Veo 3 engine and now adds audio capabilities.
It also announced Imagen 4, the company’s latest image generator. This, according to Google, is better at generating text in images, among other improvements.
Google will also make its AI Mode – a chat-like version of search – widely available in the US. This will run on a modified version of Gemini 2.5.
In addition, Google is beginning to allow users to associate personal data with large language models. The first application of this is to automatically generate personalized smart replies in Gmail, based on your e-mail history. This feature will be available to paying subscribers this summer, CEO Sundar Pichai said.
What it costs
Google also announced two new AI subscription tiers: Google AI Pro and Google AI Ultra. These replace the previous Gemini Advanced plan and offer users access to advanced AI tools and services.
Google AI Pro costs $19.99 per month and offers access to Gemini 2.5 Pro; Veo 2 for video production; 2TB of storage for Gmail, Drive and Photos; NotebookLM Plus; image to video conversion tool Whisk; and early access to Gemini in Chrome and the Flow AI filmmaking tool (up to 100 generations per month).
In addition, university students in Japan, Brazil, Indonesia and the UK will receive free access to Google AI Pro for one school year.
Google AI Ultra ($249.99 per month) includes all the features of AI Pro, with the addition of 30TB of storage for Gmail, Drive and Photos; YouTube Premium subscription; access to Veo 3 with native audio generation; expanded capabilities in Whisk and NotebookLM; access to the upcoming Gemini 2.5 Pro Deep Think mode; Agent Mode (an experimental feature for automated workflows); and Project Mariner, an AI assistant that can manage up to 10 tasks at once, such as research, bookings and purchases,
Business AM
Subscribers 0
Fans 0
Followers 0
Followers