Connect2ConnectOnline

Google Meet users now have access to Beam, Project Astra, and additional AI-powered tools.

Google introduced a number of generative AI-powered products at its yearly I/O conference for developers and other techies. These included Project Astra’s integration with Gemini Live, a 2D-to-3D video communication tool dubbed Beam, and Google Meet’s speech translation is more natural.

Later this year, Beam, a device being developed in partnership with HP, will be made available to early adopters. Using cameras and artificial intelligence to portray subjects on a 3D lightfield display, Beam uses a video model to transform 2D video streams into a realistic 3D experience, according to Google.

The new AI translation feature is important because it will try to match the speaker’s voice, tone, and expressions to make the translated interaction feel more natural. In the meantime, Google AI Pro and Ultra subscribers in beta will get near real-time translations in Google Meet for the English and Spanish languages, with more languages to follow. This year, workspace business customers will get the feature for early testing.

Additionally, Gemini Live users will be able to utilize Project Astra’s camera and screen-sharing features thanks to the integration of Google’s Project Astra AI assistant product, which is currently available for Android users and will soon be available for iOS users.

Google also said that it was adding agentic features to the Gemini app, Chrome, and Search to provide more customized and independent online search experiences.

“I’m very thrilled at how quickly the model is developing. Since our first-generation Gemini Pro model, elo scores – a gauge of progress have increased by more than 300 points. In every category, Gemini 2.5 Pro dominates the LMArena rankings today”,stated in his speech by CEO Sundar Pichai.

He also said,”Our world-class infrastructure is what makes model progress possible. Ironwood, our seventh-generation TPU, is the first developed specifically to power thinking and inferential AI workloads at scale. It offers ten times the performance of the previous generation and packs an astounding 42.5 exaflops of compute per pod”.

• The Future Of Air Travel: Why The 100ml Liquid Limit May Soon Disappear
• Rising Temperatures, Rising Tempers? The Surprising Link Between Heat And Aggression
• India’s Lost Dragonfly Returns After 110 Years
• Snapchat Introduces Safer Story Sharing For Under-16 Users
• Google Launches Gemini 3.5 Live Translate
• The New Dating Trend Everyone’s Talking About: Puffer-Fishing
• Tea, Coffee, Juice: The Worst Empty Stomach Mistakes
• Instagram Introduces Profile Grid Customization