A Tale of Two Internets

Whether we realize it or not, we’re seeing two internets emerge: one built for human audiences, and one built for AI.

These internets operate simultaneously: one optimized for rich, human reading experiences, and another with clear structure, questions, and answers, ideally in a text-only format.

We’re already seeing brands and publishers optimizing for AI search while also creating more editorial, story-driven content.

For brands, it’s crucial that marketing content needs to be findable, discoverable, and optimized for agents. Discovery no longer starts on homepages or even in search engines; it starts with AI acting on a user’s behalf. According to a March 2026 survey from G2, 51% of B2B software buyers now begin their research with an AI chatbot more often than with Google, up from 29% in April 2025. 

It’s no longer a side project, but should be built into a brand’s go-to-market initiatives. 

So what does this look like in practice? In The Economist’s case, it looks like agent-ready versions of content structured specifically for AI answer engines, already outside its paywall, such as marketing copy and B2B sales material, and complementary versions of the same content for humans.

Brands can apply the same thinking to their brand content by creating discoverable demand-generation content that lives outside the gate to be surfaced by LLMs and editorial, and by creating exclusive content written by experts that audiences must provide information (like an email) to access.

This significantly changes how we map content to the marketing funnel. The marketing funnel, as it relates to content, was built for one internet. Each piece of content was mapped to a stage, a metric, and an owner. Buyers entered at the top via SEO, whether through a blog post or a paid ad, were nurtured through demand-gen content in the middle, such as a case study or white paper, and were then handed off to sales at the bottom via a demo. 

But that model no longer works when AI becomes the first layer of discovery. In a single-path internet, you own and control the discovery. In a dual-path internet, you don’t. This creates a fork; either your buyer finds you in the AI answer and enters into a new funnel driven by ungated brand content through to expert-driven editorial, or your brand isn’t surfaced and the journey ends before it starts.

A dual-path internet has two types of purpose-built content: content built for discovery and content built for trust and engagement. 

The content that surfaces on AI answer engines gets audiences into a brand’s world, while editorial-driven content keeps them engaged by building trust and credibility. 

Agents might drive discovery, but conversion (paid memberships, subscriptions, revenue) depends on trust and engagement. Without the trust layer, the discovery layer's economics collapse.

Brands that are experimenting with this model aren’t just changing their content; they’re changing their operating model to match. The Economist is tackling this by shortening product delivery cycles, folding generative AI into the development process, and reorganizing production around small, cross-functional team “pods” that can move at AI speed. 

Current orgs aren’t set up this way. Content is spread across various functions that serve multiple audiences. What The Economist is trying to do is centralize it into one function. Brands can adopt this model, leveraging four components to make it work:

Brands still optimizing for one version of it aren't behind on a trend. They're operating on a broken assumption — that discovery still starts where they can see, track, and control it.

When AI becomes the first layer of discovery, you don’t just have a new awareness channel—discovery happens somewhere else entirely. Your buyer is already forming an opinion of your brand before they've even touched anything you've built. The content that shapes that opinion either exists and was built to travel, or it doesn't, and someone else's does.

A dual-path internet fundamentally shifts who owns your brand’s first impression — and whether your brand shows up at all. And to win, you need a structure built to run two content jobs simultaneously without losing coherence, authority, or narrative control across either.

Next week, I’m going deep into how to build out this operating model and the tactical shifts you can make to adopt it in your own org.

What’s trending across the interweb

Media news, launches, and insights

What I’m consuming this week