I spoke with Google’s Head of Android about the future of AI – and smart glasses are involved

Last week, deep in the heart of Mobile World Congress, in a remarkably refined, albeit temporary, meeting space, I couldn’t help but reflect on how far Google’s Android ecosystem had come. Five short years ago, Android’s mission was focused on personalized privacy controls and smart home integration. 

Also: Best of MWC 2025: 10 most impressive products that you might’ve missed

Now, amid towering Android statues, Gemini balloons, and miniature figures rappelling from bookshelves, the narrative had shifted decisively toward integrating AI-powered experiences into almost every aspect of the mobile experience.

At the center of this evolution sits Sameer Samat, Google’s Head of Android, who graciously carved out time during the conference chaos to discuss Gemini Live’s new real-time video and screen-sharing capabilities — features powered by DeepMind’s Project Astra — and what they reveal about Android’s AI-driven future.  

Gemini Live’s new video and screen-sharing features

Samat’s enthusiasm for Gemini Live’s new features was palpable from the moment we began discussing the updates, even if he first praised his team’s measured marketing of Circle to Search last year. “We didn’t really use the word AI once,” he explained. “It was just like, here’s what you can do with it and how it helps you.” Samat plans to continue broadcasting the benefits of new Android features in place of the simple fact that AI is driving much of it.

New to Gemini Live is the ability to point the smartphone camera at anything and immediately discuss it. At one point at the conference, I was taken to a special room in Android Avenue, where a ceramic vase sat on a table next to several glazed color samples. Gemini Live, in conversation, was able to help decide which colors would match best on the blank vase through voice queries along with live video from the camera on a Pixel 9 Pro XL.

Also: Gemini Live is expanding and can now talk about what’s on your screen

Gemini Live will also be able to recognize screen sharing content and engage in contextual conversations around what it sees. Samat envisioned someone price-hunting online for a new car and using Gemini Live to say, “Hey, what’s the average price of all the things we just looked at?” Gemini Live could then average the cost of every vehicle browsed during the session. 

Project Astra comes to the smartphone first

Crucially, these capabilities stem from DeepMind’s Project Astra, a research initiative to develop a “universal AI assistant” that simultaneously processes voice, visuals, and environmental context. Having experienced Project Astra firsthand at the Google campus last December, these new features immediately took me back to the power and possibilities I enjoyed in my demo. 

Samat admitted that while these features coming to the smartphone are exciting, it only scratches the surface of what’s possible. “Gemini with the Astra capabilities integrated with it, which is where we are now on the phone, really sort of foreshadows what’s possible on glasses.”

Imagine thumbing through an engineering textbook, Samat illustrates, and asking Gemini to remember a diagram shown on the pages. Later, once it’s time to solve a problem, you ask Gemini to refer to the knowledge from the diagram it saw earlier. “It can help you start working through problems. That’s pretty awesome.” 

Also: The best Android phones to buy in 2025

That context window of knowledge is a key component of the power and promise of these new multimodal features in Gemini Live. The trick, however, is remembering to utilize the technology in advance of that moment so it can take note of those ambient details in the first place. 

Samat agrees that this seamlessness is a big challenge for the promise of this technology right now. “The things that feel the most magic are the ones that are the most seamless,” he continued. “We’re all trying to get to that and working hard at it.”

Balancing innovation and consumer trust

Our conversation pivoted to the contrast between the industry’s big bets on artificial intelligence and consumer readiness and acceptance of it. As I roamed the halls of MWC, I made note of the strategic importance of AI being signaled by almost every company on the floor.

This contrasted with a general sentiment from attendees and consumers that these features and products aren’t nearly as desirable, useful, or secure as the industry wants them to believe. “One of the big things we’ve got to do as an industry with consumers to build trust is talk about the benefits,” Samat points out, “not just the technology.”

Samat also refers to Google’s decade of experience with augmented reality and the debut of Google Glass back in 2013, as pivotal examples of how to approach this growing technology in a way that addresses the privacy concerns of its users. “We’re trying to move in, what our CEO calls, a bold but responsible way around this,” Samat observed. “We want to advance technology, but we want to make sure that we’re doing it in a way that is thoughtful.” 

Also: AT&T just made a video call from space – and the implications are astronomical

This can sometimes mean that while Google might be poised to benefit from being an early mover, as was the case with Google Glass, the company might instead choose to take its time. “That means sometimes we won’t always be the first to do something even if we have that capability.”

Reflecting on modern smart glasses, Samat suggests that Google’s pioneering work in wearables has played a significant role in shaping the industry standard of alerting passersby when a camera on glasses is recording. 

“A lot of those experiences help us understand that you need to be transparent with people on what’s going on.” When it comes to privacy, Samat says the company continues to look deeply into keeping its users comfortable and protected with news to share soon.

Will agents eclipse apps?

The rise of AI agents on smartphones, like Google’s Gemini 2.0, which powers features like Gemini Live and Project Astra, has raised concerns among some developers that their apps may be sidelined in the march toward agentic action. 

When asked how Samat squares up Google’s progress in agents with its long-standing relationship with third-party developers, he quickly admits, “I’m an optimist on this. I think it’s going to be a partnership where these things are going to work together.”

Samat recognizes the vital service that third-party apps bring to the Android ecosystem and sees them more as an experience than a single-purpose API. He also likens AI agents to personal assistants. “If we each had an assistant,” Samat noted, “they would be using some of these services and helping you use some of these services in the right way. 

I don’t think they would be replacing those services.” In that sense, Samat sees on-device agents as ways for users to get more done rather than replacing those things entirely.

The end goal

As our discussion wound down, I asked Samat how he envisions the AI narrative evolving over the next few years. Samat made no bones about the importance of this moment in AI to Google and, more broadly, the entire industry. “AI is a super important technology,” Samat said. “It’s as big a shift as mobile, as big a shift as personal computing.” 

Also: Roll over, Darwin: How Google DeepMind’s ‘mind evolution’ could enhance AI thinking

Samat sees this moment as an opportunity to illustrate to consumers and developers what the benefits of AI are in how we use technology today.  “I think it’s incumbent upon all of us in the industry to talk more about those benefits and make those benefits a reality.”

In talking about those benefits, Samat reiterated the importance of allowing “AI” to take a backseat to what the technology can do for consumers. “Consumers just want to know if this stuff can help them actually get stuff done.” If Samat’s team succeeds, the greatest endorsement could come from users who experience the benefits of AI without even realizing it.