Kalakshar US

Google first teased Project Starline in 2021, billing it at the time as a “magic window” that uses special hardware, computer vision and machine learning to create an almost holographic video call experience. Since then, we’ve checked out an early version of the experiment for ourselves and learned last year that the company is teaming up with HP to bring a scaled-down version of the product to enterprise clients. At I/O 2025 today, Google announced that Project Starline “is evolving into an AI-first 3D video communication platform” called Beam. CEO Sundar Pichai said onstage that the first devices will be available later this year to “select customers,” though there’s no word on pricing just yet.

Adding to the computer vision and machine learning that already go into Project Starline (now Beam), Google says the platform will “use AI to enable a new generation of devices that help people make meaningful connections, no matter where they are.”

Part of this is a “state-of-the-art AI volumetric video model” and that, plus some power via Google Cloud, combined with the light field display from before, is what makes the calls “appear fully 3D from any perspective.” There are six cameras built into the Beam system to capture you from various angles, and the system uses AI to merge the streams and render you and your caller on the light field display. It’ll track your head movement to make sure it’s delivering the data to your eyes at the right angles, and do so at 60 frames per second.

Though we don’t know yet how much a unit of Google Beam might cost, it might be worth pointing out that Logitech made a different version called Project Ghost, and estimated that it will cost between $15,000 and $20,000 per booth depending on the configuration. To be clear, though, Project Ghost is not holographic or rendering 3D, it’s just a large TV streaming 2D video that’s so big it makes your callers seem life-sized.

This article originally appeared on Engadget at https://www.engadget.com/ai/goodbye-project-starline-hello-google-beam-3d-video-conferencing-171056302.html?src=rss 

If you find Google’s live translation tools useful, you’ll soon be able to use them more naturally during video calls and meetings. The company announced today that it’s bringing the feature to Google Meet. Starting this week, AI-powered, real-time Spanish translation will be available in the app. Google says more languages are on the way and that the technology is “very, very close to having a natural and free-flowing conversation.” 

During a brief demo, live translation in Meet matched the speaker’s tone and cadence, and was even able to channel expressions. There’s no doubt this will be useful for many people, especially on work calls with colleagues in other countries. Live translation will allow everyone to speak in the language they’re most comfortable with, and the technology will do the heavy lifting. Before now, you had to rely on live captions in Google Meet to do any translation, so not having to read those will allow users to be more in-tune with the conversation. 

Starting today, live translation in Google Meet will be available to Google AI Pro and Ultra plans in beta. The company says the new features will be available for enterprise users later this year as part of an “early testing” phase. 

This article originally appeared on Engadget at https://www.engadget.com/apps/google-brings-live-translation-to-meet-starting-with-spanish-174549788.html?src=rss 

Last month, Gemini Live camera and screen sharing became available on Pixel phones as part of Google’s April feature drop. But today, at Google I/O 2025, the company announced that it’s bringing that capability to all compatible Android and iOS devices as part of a free update.

Available inside the Gemini app, the ability to share your phone’s camera or content from your screen with Google’s AI is meant to make it easier and more natural to get answers about complex topics. For example, instead of describing a situation only using your voice, you could simply point your handset’s camera at something while Gemini uses object recognition to analyze the scene and do things like identify a particular species of animal or tell you want kind of screw or bolt you might need to perform a repair.

While Google doesn’t have a concrete timeline for when these features will land on your device, the company says that “in the coming weeks,” Gemini Live will get deeper integration with some of its other apps including Google Maps, Calendar, Tasks, Keep and more. The idea is that when you ask a question, Gemini will have more pertinent info to pull from so it can come up with more relevant and useful replies. Thankfully, for anyone concerned about data privacy, Google says these app connections can be easily managed inside the Gemini app and that users can adjust their settings at any time.

With Google’s continued push into AI and the expansion of Gemini’s abilities, it was only a matter of time before camera and screen sharing arrived on non-Pixel devices. That said, with the feature having just become available on first-party phones last month, it’s nice to see that there won’t be a huge wait for more folks to get access to these features.

This article originally appeared on Engadget at https://www.engadget.com/ai/screen-and-camera-sharing-in-gemini-live-is-heading-to-all-android-and-ios-devices-174547592.html?src=rss 

Like it or not, Google’s Gemini AI is worming its way into more of your Workspace apps. At I/O 2025, CEO Sundar Pichai announced that its upcoming Gmail app will have a new feature called Personalized Smart Replies. The idea is that with your permission, Gemini can examine your past emails and Google docs, then help you craft responses to personal- or business-oriented emails. 

In an example onstage at I/O, Pichai used the example of a friend taking a road trip through Colorado to Utah, a trip the friend knew Pichai had taken before. Gemini is able to look through all the notes in Drive, looking through email for related reservations and finding an itinerary in Google Docs. It’ll pull all that stuff together and then draft an email that matches the typical tone, style and even words that Pichai commonly uses. 

All these new features are coming to Google’s Workspace subscribers starting this summer. Though they may prove useful to many users, some have found Google’s addition of recent Gemini features to be intrusive and unwelcome. 

This article originally appeared on Engadget at https://www.engadget.com/apps/personalized-ai-powered-smart-replies-are-coming-to-gmail-172422777.html?src=rss 

Last year at Google I/O, one of the most interesting demos was Project Astra, an early version of a multimodal AI that could recognize your surroundings in real-time and answer questions about them conversationally. While the demo offered a glimpse into Google’s plans for more powerful AI assistants, the company was careful to note that what we saw was a “research preview.”

One year later though, Google is laying out its vision for Project Astra to one day power a version of Gemini that can act as a “universal AI assistant.” And Project Astra has gotten some important upgrades to help the company accomplish this. Google has been working on upgrading Astra’s memory — the version we saw last year could only “remember” for 30 seconds at a time — and added computer control so Astra can now take on more complex tasks.

In its latest video showcasing Astra, Google shows the assistant browsing the web and pulling out specific pieces of information necessary to complete a task (in this example, fixing a mountain bike). Astra is also able to look through past emails to find specific specs of the bike in question and call a local bike shop to inquire about a replacement part.

This story is developing, please refresh for updates.

This article originally appeared on Engadget at https://www.engadget.com/ai/project-astra-googles-vision-for-a-universal-ai-assistant-is-pulling-into-focus-174539875.html?src=rss 

Apple has sent the invites for its in-person WWDC 2025 festivities on Monday, June 9, featuring the keynote session at 1PM ET/10AM PT. Attendees will be able to watch the keynote presentation at the company’s Cupertino campus, as well as meet with developers and participating in special activities. For everyone who hasn’t received an invite to Apple Park, the keynote will stream online. Developers can also participate in the rest of WWDC’s programming online for free.

We’ve already got pretty high hopes for the keynote announcements, with a lot of potential news expected about the upcoming redesign for iOS 19. We’ve heard that the operating system could have features including AI-powered battery management and improved public Wi-Fi sign ins, and our own Nathan Ingraham has penned an impassioned plea for a normal letter “a” in the Notes app. The full WWDC conference runs from June 9-13.

This article originally appeared on Engadget at https://www.engadget.com/big-tech/apples-wwdc-2025-keynote-will-be-june-9-at-1pm-et-150621700.html?src=rss