Google Imagines a Future Where Everyone Shops in Ads
There was a festival-like quality to this year's Google I/O and GML, both held in Mountain View on back-to-back days for the first time, and that's worth dwelling on for a moment.
(Arguably too much) sun on the lawns, Gemini-driven ambient techno music between sessions, and a kind of communal exuberance pulsing through both the Shoreline Amphitheater and the Googleplex, the home of this year’s GML.
Ten Years and a Pair of Shades Later
Ten years ago, during the second-ever episode of Future Commerce, there was a breakdown of Google I/O 2016 (risky click, we were not great podcasters back then).
The big announcements that year were two chat apps, Allo and Duo, the latter featuring a "Knock Knock" preview camera that we agreed would produce a lot of accidental on-screen nose-picking.
But Allo would come with the new Google Assistant, and we spent 30 minutes dreaming about where that might lead.
We called it "the Conversational Economy," borrowing from a newly-penned Greylock essay. I sketched out a wild scenario:
“You'll be standing in a restaurant, Google will already know the menu, you’ll tap a button to chat with the business, and Maps and the Assistant will handle the rest. Payment, routing, conversation, all of it absorbed into the chat.”
We thought we were being visionary.
This week, a decade later, Nishtha Bhatia, decked out in the new Google-powered Gentle Monster intelligent eyewear, stood on the I/O stage, asked Gemini for directions to a place she had met a friend the previous week, then had it place her usual DoorDash order, pay, and tip. No menu lookup necessary, and no need to take the phone out of her pocket.
The Conversational Economy we were excited about in 2016 walked out onto a stage in Mountain View this week, ten years later, in Korean fashion.
And Allo? It was shut down in 2019.
Meanwhile, other announcements showed us that commerce is slowly permeating other parts of our online cultural experience: Universal Cart means you can add a brand’s products to a basket from Gmail, YouTube, Search, and even Ads.
You were told the Conversational Economy would arrive in a chat app. Instead, it looks like an ambient layer that runs across glasses, search, video, and email, with a persistent shopping layer that follows you across all of them.
Demis Hassabis, CEO of Google Deepmind, closed his remarks by suggesting we are “standing in the foothills of the singularity,” a phrase you do not drop casually at a developer or marketing conference.
Nick Fox, SVP of Knowledge and Information, in our conversation at GML (Apple Podcasts, Spotify), told us more than once that his team "feels the responsibility" of building products that billions of people use.
Suresh Ganapathy, who was recently coaxed away from Meta by Google to lead Product Management for Consumer Shopping, walked us through Universal Cart and AP2 on the show from the I/O grounds (Apple Podcasts, Spotify). He described tamper-proof digital mandates that create accountability among the shopper, agent, and merchant.
In 2016, Google was hoping to win your trust in a chat window you spent a minority of your time in. But today, Google is asking what it owes a society that spends the majority of its time, and therefore, sees the bulk of its reality through a pane of glass that they largely control.
The product announcements deserve a breakdown, and you'll find it below. The deeper question, and the one we'll keep returning to this year, is what we are trading in exchange for all this convenience.
Future Commerce was inside the festival tent for the first time this year, and, yes, the view was impressive.
But it was a good reminder of how long it takes for the future to actually get here.
— Phillip
‘Glasshole’ Glow-Up.
Perhaps one of the more tactically tantalizing moments at I/O was when the world got a first look at Google’s two new smartglasses. Developed with Samsung hardware and the design aesthetics of high-fashion South Korean label Gentle Monster and brand-of-the-people Warby Parker, Google is officially rebranding these products as “intelligent eyewear.” We can’t blame them for shying away from the “smartglasses” designation after the Google Glass debacle of the early 2010s, but does it warrant the glow-up?
Much has changed in the world (and the tech) since the nearly forgotten “Glassholes” era. Bygones will be bygones because the demo of the new eyewear was extremely compelling. Without pulling out her phone, Senior Product Manager Nishtha Bhatia asked for directions to the “place I met my friend Gianna last week,” added a stop for coffee, and had Gemini “put in my usual order.” In both scenarios, she didn’t provide any additional context; Gemini was able to pull the intelligence on its own. Gemini then placed the order on her behalf via DoorDash, paid, and added a tip. Gemini then shared a verbal summary from her family group chat and added details about the upcoming gathering being discussed to her calendar. She then wrapped it all up by taking a photo of the I/O audience and asked Gemini to use generative AI to turn it into a cartoon and add a blimp in the sky that read “Google I/O 2026.”
These are indeed intelligent glasses, and the two very different designs offer something for everyone: the hipsters-turned-corporate professionals (Warby Parker) and the trend-forward fashion fanatics (Gentle Monster). When combined with Google’s ever-expanding suite of AI tools and its pursuit of “personal intelligence,” there seems little doubt that these new glasses will offer a compelling window into the world and, very possibly, be the primary way we interact with it in the near future. (Unless you’ve already invested in Meta Ray-Bans, that is.)
These first iterations will be audio-only, set to be unleashed into the wild this fall, but versions with in-lens display are already in the works. They’ll pair with both Android and iOS devices out of the gate, providing interoperability and flexibility for those Apple diehards.
Invisible Signals.
We may soon no longer need to rely on our peers with the spidey sense to detect AI-generated content. Google’s SynthID, a digital watermark that is invisible to the human eye, identifies AI-created or AI-modified images and videos, such as the one that made the rounds last year of Google CEO Sundar Pichai happily consuming a McDonald’s burger at the foot of a Cybertruck alongside Elon Musk, Jensen Huang, and Sam Altman.
While that image could easily be identified as fake by anyone who knows Pichai is a vegetarian, most deepfakes aren’t as easy to spot. In fact, Pichai shared that most people can only identify high-quality deepfakes a quarter of the time. SynthID aims to change that.
The tech actually launched three years ago and has been used to watermark 100 billion+ images and videos, but as with any kind of authentication standard, it only matters if everyone buys in. This is especially true as content is seen, remixed, and shared across various platforms. The big headline from this week is that several major companies, including OpenAI, Kakao, and ElevenLab, have agreed to bring the SynthID technology to more AI-generated content.
Plus, users can now check for the SynthID watermark not just in the Gemini app, but in Search and Chrome, marking a potentially significant step in unblurring the line between what is human and what is AI.
One Cart to Rule Them All.
Mentions of the Universal Commerce Protocol (UCP) were so frequent at both I/O and GML that there were jokes that fandom tattoos might be in the cards. (We’ll leave that to the Googlers for now.)
But there’s no denying the heft of this tech. While the promise of UCP is great, it’s already delivering value in distinct use cases by creating a common language for AI agents, merchants, and brands to communicate across the entire online funnel, from discovery to purchase, and even post-purchase.
That promise gets even weightier when you consider the partners now involved in its development, with Amazon, Meta, Microsoft, Salesforce, and Stripe joining an already stacked lineup of foundational partners that includes Shopify, Target, Walmart, Etsy, Mastercard, Visa…the list goes on.
When we sat down with Ashish Gupta, one of the chief architects of UCP, he detailed the deeply layered complexity of the task at hand. Anyone who has ever been involved in building a checkout experience knows it’s not for the faint of heart; now scale that to account for every potential merchant, agent, and payment provider and scenario. It’s extremely clear why a concerted effort from so many different industry leaders, even direct competitors, is essential to the protocol’s success. As VP/GM of Ads and Commerce, Vidhya Srinivasan, said, “It may very well be the first time we’ve all agreed on something.”
Google is well on its way to wrangling that beast with the omnimodal aim of delivering live inventory and pricing accuracy at the moment of the impression, transactable at the moment of inspiration. One of the more compelling new applications, arriving soon in a phased release, is Universal Cart, which has an essence of other attempts at cross-channel, multi-brand checkout (namely, Buy on Amazon), but with a broader reach that encompasses Search and Gemini interfaces, as well as Gmail and YouTube.
Yet Universal Cart isn’t relegated to a platform; it lives inside the internet, riding the consumer behaviors Google already owns. Once a product is added, Gemini works in the background, finding deals, monitoring price drops, offering insights on price history, and alerting users
when an item is back in stock. When ready to buy, shoppers can check out directly with Google Pay or transfer the items to the merchant’s site to complete the purchase there. Either way, the brand remains the merchant of record, reaffirming that disintermediation has been a myth all along.
Creator Paradise or Generative Wasteland?
There may be no accounting for taste, but there’s also no denying that today’s cultural tastemakers are the fleet of creators that fill our feeds. And in the world of generative engine optimization (GEO), where organic rank is driven by experience and expertise, much of the relevance value signaled to LLMs comes from online commentary from creators and customers.
Brands’ advertising aim is no longer to show up when called by a keyword, but rather to answer relevant questions. During GML, Ulta Beauty’s Head of Agentic, Augustina Sartori, described this shift as creating “conversational experiences,” and it explains why creators have never been more important in the marketing ecosystem. In fact, Katie Babcock, Director of Performance Solutions at Google, noted that YouTube mentions are now one of the highest contributors to AI visibility. (Eat your heart out, Reddit.)
But it’s clear the picture isn't so black-and-white. As SVP of Knowledge and Information, Nick Fox was preaching the expanded opportunities for creators on stage, his colleagues outside were demonstrating how quickly the company’s AI creative studio, Flow, could generate custom UGC. The final outputs are scary-realistic and can be generated from a single image of any human, or, for that matter, any AI-generated human amalgamation.
The Pursuit of Personal Intelligence.
All the years we spent being intermittently creeped out by overly prescient targeted advertising may have only been training for this moment. Digital clairvoyance is now the feature, not a flaw, and Google is betting that users will start to buy in.
Earlier this year, Google rolled out “personal intelligence” in AI mode, wherein users opt in to connect Google apps like Gmail, Photos, and Calendar to Gemini, which then uses all of that data and knowledge to help with everything from suggesting ways to kill the time during a layover to getting your dogs registered at a new kennel.
Now, with the just-announced AI assistant, Gemini Spark, Google is explicitly aiming to become users’ “active partner that does real work on your behalf and under your direction,” Josh Woodward, VP of Google Labs, Gemini app, and AI Studio, wrote in a company blog. Spark will work in the background to handle recurring or triggered tasks and create and handle complex workflows. This is the ultimate vision, channeling the ubiquity of Google’s presence in our lives into irreplaceability.
Google’s not alone in pursuing this aim, with Meta also actively building toward its own twist on the same idea: “personal superintelligence.” These platforms are already deeply interwoven into our lives. If all that activity and knowledge is to be funneled through and managed by their AI agent, well, that’s a relationship that would be near impossible to sever. If we get personalized utility and a lighter workload, they’ll get the ultimate reassurance of lifelong loyalty. Talk about a trade-off.
.svg)
