Add custom AI agents to your marketing team with Agent Cloud
For nearly a decade, the QR code has been the undisputed, if clunky, bridge between physical media and digital experiences. It’s made it totally normal for global audiences to point their smartphones at surfaces to unlock content.
Yet, as a creative element, the QR code is an eyesore. It’s a jarring interruption to any brand’s visual storytelling efforts. For many, it’s an unwelcome reminder of a pandemic era in which dining out meant scanning a QR code simply to access a restaurant menu.
But what if the omnipresent QR code is reaching the end of its lifespan? If a recent conversation I had with Matty Beckerman, CEO and founder of IRCODE, is any guide, the future of interactive media might involve ditching this wonky black-and-white matrix.
Beckerman’s big idea relies on computer vision AI that treats the creative asset itself as the “destination.” But before we dive into all of that, a brief reminder of where the QR code Beckerman aims to replace actually came from.
To understand where visual commerce is heading, it helps to understand where QR technology started.
The QR (Quick Response) code was invented in 1994 by Masahiro Hara, an engineer at the Japanese automotive company Denso (now Denso Wave). Originally, it was developed to solve a specific manufacturing problem. Standard one-dimensional barcodes possessed limited storage capacity, which forced workers to scan multiple barcodes just to track a single car part.
Hara's two-dimensional matrix solved this by holding substantially more data and scanning rapidly from multiple angles. In a vital strategic move, Denso Wave made the QR specifications freely available to the public, laying the essential groundwork for global adoption.
While QR codes saw steady consumer growth with the advent of camera-equipped smartphones, their usage exploded during the COVID pandemic. The sudden demand for contactless solutions led to the rapid, widespread adoption of digital restaurant menus, touchless payments, and virtual check-ins.
This shift fundamentally altered consumer behavior in ways that, while radical at the time, simply feel normal now. By 2022, 89 million people in the United States alone scanned a QR code with their mobile devices, representing a massive 26 percent increase compared to 2020.
After that massive burst in QR code adoption, it’s possible that the world (and the technology) is ready to move on.
During the Cannes Lions Festival of Creativity and Sport Beach, I sat down with Matty Beckerman, CEO and Founder of IRCODE (Image Recognition Code) to discuss how spatial computing and advanced AI are offering alternatives.
IRCODE utilizes patented computer vision to scan live broadcasts, streaming video, and static images, surfacing deep data layers without a QR code in sight.
One issue they’re solving for is an aesthetic one. As Beckerman bluntly puts it, the visual compromise of the QR code is a major pain point for creators, marketers, and anyone else putting creative assets into the world. ("I don't think anybody wants to put that ugly black and white QR code on a piece of artwork that you've created,” he said.)
While tech giants have built proprietary visual search engines—such as Google Lens, Amazon Lens, and Pinterest's visual tools—IRCODE's strategy is inherently friendlier to brands and broadcasters.
They build white-labeled lenses directly into a company’s native application. This isn't a future proof-of-concept, it’s happening right now. IRCODE recently rolled out the "KSL Lens" with NBC affiliate KSL in Salt Lake City, turning the station’s local broadcast entirely interactive.
More notably, Sinclair Broadcasting Group—which controls roughly 40% of the U.S. television affiliate market—recently announced a strategic investment in IRCODE. The roll-out begins with two stations this summer, expanding to 12 more before the end of 2026.
For television networks, streaming platforms, and major brands, the financial implications of this shift are massive.
IRCODE strips away the traditional barcode entirely, relying instead on patented computer vision and video recognition. The system is anchored by a massive master database where broadcasters and brands register their visual assets, attaching specific metadata and digital destinations to those images. Because the AI cross-references the live image against this deterministic, locked database rather than scraping the open web, it bypasses the risk of AI hallucination and protects brand intellectual property.
For the consumer, this creates a frictionless workflow embedded directly into the native applications they already use.
Consider the classic "second-screen" problem: a viewer is watching a reality television show and spots a dress they want to purchase. Today, they leave the broadcast, search the web for 10 minutes, and frequently land on a competitor's knock-off.
With native image recognition, that same viewer simply opens the network's app and points their camera at the TV (or utilizes IRCODE's "Tap to Scan" feature if they are already streaming on their mobile device).
The AI instantly matches the dress to the database and surfaces a direct, authentic purchase link. The transaction is completed in seconds, ensuring the network and the brand retain the first-party data, the direct consumer engagement, and the final conversion.
By controlling the lens, the brand keeps the user within its own ecosystem. And if a consumer scans a product via a broadcaster's app, the network retains control over the transaction, rather than letting Google or Amazon take a cut.
The underlying architecture of IRCODE is rooted in deep technical engineering. The company's chief technology officer, Philip Holsteiner, originally invented the core patent to allow autonomous drones to fly over cities and navigate precisely without relying on GPS, attaching localized data directly to real-time imagery.
As I just mentioned, adapting this high-fidelity video recognition to consumer media requires building a massive master database of verified images embedded with rich metadata.
In an era dominated by large language models, this database provides an invaluable guardrail against hallucination. Instead of an AI model training on open-source web data or Wikipedia to guess what it is looking at, it cross-references IRCODE's locked database to identify the precise asset.
Furthermore, this structural layer introduces a critical element of intellectual property protection. If an item is registered within the IRCODE system, its metadata can be blocked from unauthorized scraping and AI model training.
It moves the technology from guessing generic objects to mapping deterministic commercial pathways. That means moving from the AI saying “yup, that’s a microphone on screen” and toward identification of the specific, authentic brand model of that microphone and where to purchase it instantly.
Marketing leaders need to stop treating digital engagement as an afterthought, something slapped onto a finished commercial in post-production.
Instead, the industry must design with a digital layer from the ground up, and new tech like IRCODE is suddenly opening the door to creative solutions.
Suddenly, pretty much any surface can become an interactive portal, whether it is a live sports broadcast, a streaming phone screen, or a physical out-of-home placement. And this happens without the annoying distraction and visual noise of the QR code.
The current phase of this technology focuses on deploying custom lenses into brand apps, but the final destination is much more frictionless.
Beckerman points out a historical parallel that mirrors exactly where image recognition is headed. "If you remember pre-COVID, you had to download a third-party app in order to scan a QR code, and then eventually it went native in the camera,” he explained. “That's the future for us. We want to be native in the iPhone, native in the Android. It just works straight from your camera."
Until native camera integration becomes standard, the mandate for forward-thinking agencies and enterprises is clear: Stop relying on the visual compromises of yesterday's technology. By building creative assets with native AI image recognition in mind, it’s possible to reclaim data, protect brand IP, and offer audiences a direct path from inspiration to interaction.
