Google has hyped its project Astra up as the next generation of AI for months. It set some high expectations when 60 minutes Sent Scott Pelley to experiment with Project Astra Tools provided by Google Deepmind.
He was impressed with how articulated, attentive and insightful AI turned out to be throughout his test, especially when AI not only recognized Edward Hopper’s moody painting “Automat”, but also read into the woman’s body language and spun a fictional vignette about her life.
All this through a pair of smart glasses that hardly seemed different from a pair without AI built -in. The glasses act as a delivery system for an AI that sees, hears and can understand the world around you. It could set the stage for a new smart wearables race, but it’s just one of many things we learned under the Segment of Project Astra and Google’s plans for AI.
Look at
Astra’s understanding
Of course, we have to start with what we now know about Astra. First, the AI assistant continuously processes video and audio from connected cameras and microphones in the environment. AI does not just identify objects or transcribe text; It also claims to spot and explain emotional tone, extrapolate context and have a conversation on the subject even when you break for thought or talk to another.
Under the demo, Pelley Astra asked what he was watching. It immediately identified Kulfallsgård, a retail complex in King’s Cross and offered background information without missing out on a beat. When it appears a painting, it didn’t stop by “It’s a woman in a cafe.” It said she looked “contemplative”. “And when it was pushed, it gave her a name and a back story.
According to Deepmind CEO Demis Hassabis, the assistant’s understanding of the real world is promoting even faster than he expected, noting that it is better to make sense of the physical world than the engineers thought it would be at this time.
VEO 2 views
But Astra doesn’t just look passively. Deepmind has also been busy teaching AI to generate photorealistic images and video. The engineers described how their video models two years ago struggled to understand that the legs are attached to dogs. Now they showed how VEO 2 can conjure up a flying dog with flapping wings.
The consequences for visual storytelling, filming, advertising and yes, augmented reality glasses, are deep. Imagine that your glasses not only tell you which building you are looking at, but also visualization of what it looked like a century ago, reproduced in high definition and seamlessly integrated into the current vision.
Genie 2
And then there is Genie 2, Deepmind’s new world modeling system. If Astra understands the world as it exists, genie builds worlds that don’t. It takes a still image and makes it an explorable environment that is visible through the smart glasses.
Go forward and Genie invents what lies around the corner. Turn left and it populates the unseen walls. Under the demo, a waterfall photo for a playable video game level, dynamically generated, which Pelley was explored.
Deepmind is already using genie-generated spaces to train other AIS. Genie can help these navigate a world composed of another AI and also in real time. One system dreams, another teacher. The kind of simulation loop has huge consequences for robotics.
In the real world, robots have to fumble their way through trial and mistake. But in a synthetic world, they can train endlessly without breaking furniture or risking litigation.
Astra eyes
Google is trying to get Astra-style perception in your hands (or on your face) as soon as possible, even if it means giving it away.
Just weeks after the launch of Gemini’s screen sharing and live camera features such as a Premium Frynsegod, they turned the course and did it free of charge for all Android users. It wasn’t a random generosity. By getting as many people as possible to point their cameras to the world and chatting with Gemini, Google gets a flood of training data and real -time user feedback.
There is already a small group of people wearing Astra-powered glasses around the world. The hardware allegedly uses microled screens to project captions in one eye and delivers sound through small directional speakers near the temples. Compared to the awkward sci-fi visor of the original glass, this feels like a step forward.
Of course, there are problems with privacy, latency, battery life and the non-so-small question of whether society is ready for people walking around with semi-riscient glasses without mocking them mercilessly.
Whether Google can make the magic feel ethical, non-invasive and stylish enough to go mainstream is still up in the air. But the feeling of 2025, when the year smart glasses go, seems mainstream more accurate than ever.
