The fallen state of our Internet is a direct, if unintentional, consequence of choosing advertising as the default model to support online content and services.

Through successive rounds of innovation and investor storytime, we’ve trained Internet users to expect that everything they say and do online will be aggregated into profiles (which they cannot review, challenge, or change) that shape both what ads and what content they see.

I think the original sin was we couldn’t actually build economics, which is to say money, into the core of the internet and so therefore advertising became the primary business model…



We tried very hard to build payments into the browser.

Economics are not only built through direct money, but the symbols of credibility, trust, influence, and power that translate to the real world. These things can and are monetized, but don't have to take the form we settled on with the current quantity over quality model.

The original web was the human web, and advertising was and is one of the best possible ways to monetize the only scarce resource in digital: human attention.

If you have no guardrails around this you are *guaranteed* to trigger a race to the bottom that eviscerates the value of the information and experience of the system itself.

Subscriptions do work at smaller scales, because they are ultimately not about paying for content, but giving money to another human (or human institution); from The Local News Business Model:

Rather, a subscriber is paying for the regular delivery of well-defined value.



Each of those words is meaningful:




Paying: A subscription is an ongoing commitment to the production of content, not a one-off payment for one piece of content that catches the eye.



Regular Delivery: A subscriber does not need to depend on the random discovery of content; said content can be delivered to the subscriber directly, whether that be email, a bookmark, or an app.



Well-defined Value: A subscriber needs to know what they are paying for, and it needs to be worth it.




This last point is at the crux of why many ad-based newspapers will find it all but impossible to switch to a real subscription business model.

When asking people to pay, quality matters far more than quantity, and the ratio matters: a publication with 1 valuable article a day about a well-defined topic will more easily earn subscriptions than one with 3 valuable articles and 20 worthless ones covering a variety of subjects.

Yet all too many local newspapers, built for an ad-based business model that calls for daily content to wrap around ads, spend their limited resources churning out daily filler even though those ads no longer exist.

First, that consumer engagement has shifted into apps and walled gardens irreversibly.

And second, that Google was keeping the open web on life support, and the open web’s demise will be hastened when Google no longer has an incentive to support it.

So you need protocols, things like MCP and A2A and things that likely are going to be emerging over the coming year that will help connect in an open, reliable, interoperable way the agents that you are writing and agents that are being used so actively now by hundreds of millions of people to be able to go access content, to access services, to take action on behalf of users in fulfilling the tasks that have been delegated to them.

One aspect of this vision of the agentic web was Microsoft’s commitment to the Model Context Protocol created by Anthropic; Scott told Nilay Patel in an excellent interview in The Verge that while MCP wasn’t exactly what he would have designed from scratch, ubiquity is more important than semantic differences, particularly when you’re trying to create HTTP for AI agents.

The second part of Scott’s vision was something Microsoft created called NLWeb, a natural language interface for websites that makes them more directly accessible for agents:

As more and more of the answers appear directly, particularly in AI-based search products, traffic to websites has generally dropped.

What’s going to replace that in the agentic era, where we’ve created new schema for agents to come and talk to my website and receive some answers?

What’s going to make that worth it?

But if I make a website, I open myself up to distribution on different surfaces.

What I will get in return for that is not necessarily money - almost in every case, not money.

What I’ll get is visitors to my website, and then I’ll monetize them however I choose to: selling a subscription, display ads, whatever it is.

The problem, as I see it, is that the traffic to the web is in precipitous decline as Google referrals go into decline.

How do you fix that problem so that everyone is incentivized to keep building on the web?



I don’t know, honestly.

Unlike Bitcoin, stablecoins do not have intrinsic value downstream from a network effect, and unlike Ethereum, they are not set up to execute smart contracts or other applications; rather, their value is right there in the name: they are stable representations of value - usually the U.S. dollar…What remains is a synthetic currency that is digital but scarce, with all of the affordances that allows for, including the ability to move money frictionlessly (thus Collison’s analogy).

The analogy I think of is to the Internet itself:




Physical goods are scarce, but while you can scale up from hand-delivery, you still have to pay a lot for a delivery service, and if you cross borders you have to deal with customs.



Information used to be delivered in person, then via physical media like letters or newspapers, but now it is purely digital and freely distributed and duplicated all over the world; it is abundant.




Dollars right now are more akin to physical goods than they are to information: you can deliver it by hand, or even via delivery services like ACH or SWIFT, but the need for verification and confirmation flows introduce a ton of friction.

Stablecoins solve these problems: you can transfer them like information, while preserving scarcity, while blockchains provide verification and confirmation that scales from the smallest transactions to the biggest; meanwhile, they also earn a return while at rest thanks to the assets backing them.

In fact, we already have an excellent example of (deterministic) agents making micro-transactions at scale: the entire digital ads ecosystem!

Every time a human loads a webpage, an awe-inspiring amount of computation and communication happens in milliseconds, as an auction is run to fill the inventory on that page with an ad that is likely to appeal to the human.

These micro-transactions are only worth fractions of a penny, but the aggregate volume of them drives trillions of dollars worth of value.

The problem, as both I and Patel noted, is that this ecosystem depends on humans seeing those webpages, not impersonal agents impervious to advertising, which destroys the economics of ad-supported content sites, which, in the long run, dries up the supply of new content for AI.

What is possible - not probable, but at least possible - is to in the long run build an entirely new marketplace for content that results in a new win-win-win equilibrium.

First, the protocol layer should have a mechanism for payments via digital currency, i.e. stablecoins.

Second, AI providers like ChatGPT should build an auction mechanism that pays out content sources based on the frequency with which they are cited in AI answers.

These companies are already struggling to defend ROI — and now they're going to need to pay out to all the sources cited? How can they justify this except through — you guessed it — copying SEO and the ad model. Is this Groundhog Day?

The result would be a new universe of creators who will be incentivized to produce high quality content that is more likely to be useful to AI, competing in a marketplace a la the open web; indeed, this would be the new open web, but one that operates at even greater scale than the current web given the fact that human attention is a scarce resource, while the number of potential agents is infinite.

This is where Scott’s exhortation of openness is spot on: a world of one dominant AI making business development deals with a few blessed content creators, and scraping the carcass of what remains on the web for everything else, is a far less interesting one than one driven by marketplaces, auctions, and aligned incentives.