initial details

Google showed off a prototype of its Project Astra AR glasses, which are apparently already being tested outside the lab in the real world. The glasses are part of the company’s long-term plan to create an entire ecosystem of different hardware that supports augmented reality and multi-modal AI capabilities. According to Channel 24, at this stage they are not yet ready for release and do not offer all the features that developers dream of, so for now we will only see demo versions designed to attract the attention of consumers, developers and even competitors. to TechCrunch.

In addition to the glasses, Google also announced its new operating system Android XR (augmented reality that will combine virtual and augmented reality). It will allow developers to create new features for XR devices, such as glasses and headphones. Project Astra glasses will work on this operating system, and Samsung’s recently announced Project Moohan headset, which will be on sale next year, will also work.

Unlike the same Samsung or Apple, the Google company is not trying to close the middle stage of technology by releasing a bulky headset like Vision Pro or Project Moohan in order to someday transform it into a more convenient form factor. It directly targets regular glasses, so you can wear them at any time, even outside.

First of all, the glasses are designed to provide information, but they are not a full-fledged computer, as competitors are trying to do. It looks like the company doesn’t expect you to surf the web, play games, or watch movies. But built-in screens will give you enough information about navigation, communication, finding information, tips and so on.

We see a few usage examples in the published video:

• In the first case, the user walks down the street. He can’t use his smartphone because his hands are busy. But he doesn’t need to have his hands free to use Project Astra, as the glasses respond to his voice. He asks Gemini to tell him what happened in the conversation he had with his friends. After this we see the picture from the girl’s eyes. A short report from a summary of the last dialogue appears on the screen in front of him. Apparently, in the headset everything is transcribed in audio format. AI reported that his friends invited him to the institution for the holidays.
• The next step is to ask if the restaurant has a vegan menu. In this case, the AI ​​consults information on the Internet to find out the details about the place and displays details such as the restaurant’s rating, a review about it on the screen and itself voices the answer about the availability of vegan dishes.
• The user then asks Gemini to tell her friends that she is coming and mention the blue outfit she will choose for the evening. It’s not entirely clear exactly how this works. The glasses will probably have a function that will allow you to mix and match your wardrobe, offer new combination ideas, and require you to show them everything in your closet. So be it, but after this request Gemini creates a new chat message and sends it after confirmation.
• After that, the owner of the glasses asks him to recommend a store to buy a gift card. The AI ​​searches Google Maps, recommends a store, and once approved, displays a small map in front of the screen showing the route. So in this case glasses act as navigators.
• On her way to the store, the girl meets a street food establishment. But the entire menu has only Korean text. It was enough for the author to say “Hello, Gemini” and the on-screen system immediately understands what he is looking at and offers to translate the menu. Once approved, the translation appears before your eyes. The user asks additional questions about the dishes in order to choose the meal he likes. The AI ​​then helps communicate with the employee by translating the sentences.
• Finally, in the last demonstration, the author asks for help placing the shelf on the wall. AI suggests different options for what it might look like in terms of design. The user chooses one of them and Gemini announces the strategy step by step. At one point, he suggests making sure the board is hanging flat, so he reminds you to use a level (a device with air bubbles inside that indicates whether the surface is flat). However, the owner of the glasses does not remember where they are. That’s why he asks where the item is. Gemini replies: “I saw the last level on the dresser next door.” We have seen a similar use before when we first saw these glasses in the frame. The use of artificial intelligence was then demonstrated using a smartphone camera as an example. But it is worth considering that such an application will be possible only if you constantly show the surrounding space to your glasses.

Watch a demonstration of the glasses’ capabilities: video

Journalists who saw the glasses during the demonstration said that they had the glasses USB-C portsIncluding transparent plastic ones, under which all components can be seen. However, no one was allowed to take photos.

9to5Google’s Abner Lee says Project Astra uses microLED displays From Raxium, which Google acquired in 2022. It’s based on something unique that’s still “in research and development.”

There is a touch panel on the right frame of the glasses. One button is located on the top edge and the other button is located at the bottom. The top button takes a photo and the second button activates Gemini. Gemini remains active when turned on, but you can tap the glasses to pause the assistant.

Abner Lee said he received a demo of real-time translation and that it was “excellent.” Gemini can also translate zodiac signs. He also showed the AI ​​an album cover and asked it to play a song. AI immediately released the song on YouTube Music.

It is currently unknown when these glasses may be released. So is their price.