Open Agent Commerce: The End of the Advertising Era
TechFlow Selected TechFlow Selected
Open Agent Commerce: The End of the Advertising Era
Welcome to the New Era of Open Agent Commerce.
Navigating Web3 tides with focused insightsBy Sam Ragsdale
Translated by Chopper, Foresight News
The era of agent-powered commerce has arrived. The ACP and UCP protocols promise one-click checkout inside ChatGPT and Gemini. Soon, hundreds of millions of global consumers will effortlessly discover better products, merchants will see significantly improved conversion rates, and platforms will collect 5–10% commissions.
Yet built-in checkout in ChatGPT is merely an incremental upgrade—it cannot reshape society the way the internet did in the early 2000s. Only open agent commerce can achieve that.
Why Walled Gardens Always Fall to Open Protocols
To understand why, we must rewind to the 1990s.
Back then, two fundamentally different, competing “internet paradigms” existed.
AOL’s model: a unified monthly subscription bundling email, weather forecasts, curated content—and eventually, the entire Time Warner copyright library.
The open-protocol model: HTTP, DNS, HTML—and a browser named Mosaic.
At the time, Mosaic seemed absurd: there were barely any websites, no search was needed, and alphabetical indexes sufficed. Eight years later, AOL merged with Time Warner for $35 billion, and markets widely assumed curated content was the future.
But soon after, Mosaic and open protocols triumphed—and human civilization officially entered the digital age.
Why? Consider what would have happened had the walled garden won.
In 2004, Mark Zuckerberg would have needed to negotiate distribution terms with AOL before launching Facebook. Two Stanford students wanting to build a web index would have required CompuServe’s permission. Someone aiming to sell books online from their garage would have had to pitch their idea to Microsoft’s MSN content team.
They’d simply have been dismissed: “Go back to school, kids.” Had that happened, today’s familiar digital economy would never have existed.
The core strength of open protocols lies in having no gatekeepers. Anyone with a server and a domain name can reach the entire network. Innovation emerges at the edges—centers can’t keep up—and this dynamic ultimately sparked one of history’s largest wealth-creation waves. It reflects capitalism’s foundational logic: disruption always originates at the periphery.
Return to 1997. Tim Berners-Lee, Marc Andreessen, and others were developing these protocols and browsers. Back then, setting up a server cost tens of thousands of dollars; nobody understood why content servers should respond to requests from strangers—doing so was expensive and offered no clear return.
They designed an HTTP status code called “402”: servers could reply to users, “Pay to access this content.” But digital micropayments were utterly infeasible then—PayPal didn’t exist, and fixed credit-card fees ran dozens of cents per transaction, far too high for penny-scale payments.
Nonetheless, the internet still rose.
Google found an alternative business model: advertising. In traditional media, creators formed direct economic relationships with users. Google adopted a broadcast-economics logic, inserting third-party advertisers who paid to sustain both content supply and user reach.
Brilliantly elegant. Creators no longer needed to cultivate user relationships—they monetized attention alone. Google sat squarely between advertisers and creators, freely extracting commissions from this flow.
Micropayment demand was shelved. Open-source software flourished. The cloud revolution exploded. Server hosting costs plummeted a hundredfold. Google became the most ardent champion of a free, open internet: the more people searched, the more Google earned. To that end, it invested hundreds of billions of dollars to make the web faster, cheaper, and ubiquitous.
AI Agents End Traditional Advertising—and Walled Gardens
Fast-forward to the 2010s: industry stagnation set in.
Interest rates kept falling. Capital grew conservative. Foundational internet innovation lost its edge. Major walled gardens tightened their grip on users and stockpiled strength.
Then, in 2022, ChatGPT launched—and the world reset. Large language models (LLMs) do more than generate outputs; they synthesize information from multiple sources into concise summaries—often without directly scraping original content.
By GPT-4, the trend was unmistakable: agents would become the next core paradigm. They operate computers as skillfully as humans—but at lower cost and higher efficiency.
At this very moment, the internet’s foundational economic logic has been rewritten.
From 1997 to 2024, internet monetization centered on attention: humans browsing web pages are easily distracted by ads, and platforms monetize fragmented attention. But LLMs and agents cannot be distracted.
An ironic reversal: advertising birthed the free, open internet; vast troves of web text trained LLMs—and now LLMs, in turn, are ending advertising.
After GPT-4 launched, Stack Overflow’s traffic plunged 75%; tech news sites saw 60% drops. Tech users were early adopters—the shift will inevitably sweep across every information context online.
Built-in checkout in ChatGPT is irrelevant. The internet is humanity’s public square—and old commercial contracts have expired.
A few corners of the internet still resist Google’s crawlers via differentiated content—the classic walled gardens: Facebook, TikTok, LinkedIn. Thousands of highly paid engineers guard them around the clock, blocking bots.
But now, those walls have collapsed. AI agents flawlessly mimic human browsing behavior—fooling every defense mechanism. Over the next decade, countless gimmicky “solutions” will emerge claiming to fix this, and VCs will pour money into them—but no truly effective remedy exists. Just as fighter planes rendered even the strongest ground fortresses obsolete, so too have AI agents dismantled the walled garden.
The Open Agent Era
So what comes next? Open agent commerce.
ChatGPT’s built-in checkout is the AOL of the agent-commerce era: curated catalogs, closed loops, optimized experiences. For merchants to sell through it, they need months of business development, rigorous legal documentation, concrete five-year roadmaps, substantial revenue, massive user bases—and a story spectacular enough for the front page of The New York Times to thrill shareholders.
Open agent commerce is today’s HTTP: a minimal, universal standard enabling agents to pay on-demand for anything—data, cloud hosting, communication services, and countless yet-unimagined new use cases.
Two pioneers have already launched: Coinbase introduced the x402 protocol; Tempo partnered with Stripe to launch the mpp protocol. Twenty-eight years after the “402” status code was conceived, viable implementations have finally arrived. Modern blockchain stablecoins charge less than one cent per transaction—solving the exact cost barrier that killed micropayments in 1997.
Agents restricted to purchasing only from pre-approved white-listed merchants resemble ordinary employees holding corporate cards limited to three vendors. Agents connected to open protocols are entrepreneurs holding fully autonomous bank accounts.
No negotiations. No whitelisting. Just simple, permissionless standards.
These protocols focus on two things:
- On the agent side: How does payment happen seamlessly?
- On the merchant side: How is payment receipt verified?
LLMs excel at invoking tools they’ve never seen before. Starting with Claude 4.5+ and Codex 5.2+, agents can discover APIs, read their schemas, and use them correctly—without prior training.
Current discussions center on “skills.” These are essentially natural-language-programmable modules, freely composable. Non-technical founders can trigger full programs using everyday language:
- Order pizza from a nearby highly-rated pizzeria and check delivery status every 10 minutes.
- Turn on the porch light when the driver is 5 minutes away.
- Tip the driver $5 if delivery arrives within 30 minutes.
No coding. No programming expertise. Agents parse intent, generate native programs on-the-fly, execute them, then destroy them. Programming is no longer a specialized skill—proficiency in natural language suffices.
Skills work—but they’re transitional artifacts, the first intuitive manifestation after humans realized agents could invoke unfamiliar tools. Skills require dedicated authoring, publishing, security auditing, iterative updates—and must be pre-installed and loaded. They’re cumbersome and inefficient.
The skills craze obscures a deeper disruption: agents can combine capabilities in unprecedented ways.
Pizza is just a simple example. Real-world business scenarios are far more transformative. A small-business supply-chain agent detects a 15% price hike from its packaging supplier due to tariffs, autonomously identifies three local alternatives, requests samples, negotiates bulk pricing, and switches vendors—all before the CEO finishes their morning jog.
No API integrations. No procurement teams. No bidding processes. Just an account balance and open protocols.
Agents can pay—and create—but right now, they still can’t find what they need.
What remains is discovery: for brokers, “How do I find what to buy?”; for merchants, “How do I present my service to brokers?”
The industry has already built a universal discovery-and-registration ecosystem. Service providers simply register their nodes at x402scan.com or mppscan.com—and instantly plug into the global network of interoperable, open-protocol-compliant agents, enabling standardized supply-demand matching and automated micro-payments.
In 1997, the internet lacked a business model—and nobody grasped why servers would respond to strangers. Open protocols and advertising ingeniously solved that—and ushered civilization into the digital age.
In 2026, advertising—the stopgap solution—will fade. Open protocols and a status code buried for 28 years will replace it.
Welcome to the new era of open agent commerce.
Join TechFlow official community to stay tuned
Telegram:https://t.me/TechFlowDaily
X (Twitter):https://x.com/TechFlowPost
X (Twitter) EN:https://x.com/BlockFlow_News
a16z
