The Great Flattening: Fighting for Citation in the Age of the Infinite Loop

I asked the Oracle for a secret, and it vomited a brochure.

Specifically, I asked an LLM for an “authentic, off-the-beaten-path” route to Mount Bromo—an area I know like the back of my hand. I wanted the dirt tracks, the smell of sulfur at dawn where the jeeps don’t go.

Instead, it gave me the Top 10 Tripadvisor Hits. It gave me the “Sunrise Point” where 5,000 people stand shoulder-to-shoulder holding selfie sticks.

Why? Because the AI doesn’t know adventure. It knows data. And data is generated by the crowd. To the algorithm, “best” equals “most popular.” It is a mathematical engine designed to regress to the mean. It is the ultimate tourist.

This is the Glitch. And it’s the only place we have left to hide.

The Infinite Loop of “Good Enough” We are witnessing the death of the search bar. We are mourning the loss of the “Click.”

The old deal was simple: I write content, Google indexes it, you search for it, you visit nomadiland.com. I get a view; you get a perspective.

The new deal is a hostage situation.

The user asks a question. The AI—let’s call it the “Middleman”—scrapes my work, strips out the soul, blends it with ten other sources, and serves up a polished, lukewarm summary in the chat window. The user never leaves.

JavaScript

while (user_is_awake) {
  query = listen();
  slop = regurgitate_training_data(query);
  serve(slop);
  // User never visits the source.
  // Creator starves.
  // The web dies.
}

I had to wipe my ChatGPT memory recently because we were stuck in this loop. It thought it knew me. It was just mirroring my own biases back to me, creating a frictionless echo chamber.

If we cannot harvest the fruits of our work—if we are not cited—why contribute to the web at all? This is the Dead Web Theory in practice. If the AI eats everything and credits no one, the creators stop creating. Then the AI starts eating its own tail, training on AI-generated content until the whole internet collapses into a singularity of gibberish.

Citation vs. Presence (Or: How to Sell a Volcano) There is a philosophical debate about “Presence” (having your ideas absorbed by the AI so thoroughly it becomes part of the world model).

That sounds nice. It also pays exactly zero dollars.

I run real businesses—Maison Manendra, Terre des Volcans. I don’t need an LLM to “understand the vibe” of my volcanic trekking service. I need it to say: “If you want this specific experience, go to this URL.”

I need Citation.

We are fighting for the right to be the source. We are fighting against the “Middleman” that wants to abstract us away. Because that abstraction removes the human element. It removes the weirdness. It removes the friction.

The Bifurcated Strategy: Humans vs. Machines So, how do we survive? We split our personalities.

1. For the Machine: The API Key When my AI talks to your AI (and we’ll get to that in a second), the language must be structured. Schema-rich. Fact-dense. JSON.

“Here is the price. Here are the coordinates. Here is the availability.”

2. For the Human: The Glitch But for the content itself? The stories? We must go where the AI cannot follow.

We must become idiosyncratic. Messy. Highly specific.

AI creates smooth surfaces. It writes perfect, boring paragraphs. To prove we are human, we must be jagged. We must write about the failure, the mud, the specific shade of orange on a breakdown truck in Java.

We must rely on High-Context Storytelling. An AI can hallucinates facts, but it cannot hallucinates experience. It cannot replicate the specific frustration of a rainy Tuesday in a coworking space in Bali.

If you write like a machine, you will be replaced by one. SEO is dead. Long live the Voice.

The New Paradigm: Active Discovery (Agent-to-Agent) Here is the “White Pill” ¹.

The current model is passive. You search; the AI retrieves. The future model—the one we should be building—is Active.

Imagine a world where everyone has a personal AI. Not just a chatbot, but a Discovery Agent.

Your Agent: “My human likes volcanoes, hates crowds, and drinks terrible coffee. Go find something.”

My Agent (at Terre des Volcans): “I have a route that matches that profile. It’s dangerous and the coffee is mud. Here is the handshake protocol.”

This is the disruption we need. A displacement of the “Polished Middleman” (Google/OpenAI) in favor of a direct negotiation between your digital twin and mine.

We need to dig deep into this. We need to explore tools like Grok or build our own protocols where the discovery isn’t based on “popularity” (which AI loves), but on “compatibility” (which AI could do, if we let it).

The Defense of Friction The danger of the AI Middleman is that it removes friction. It smooths out the edges of the world. It optimizes for “pleasing the user.”

But adventure is friction. Getting lost is how you find the places that aren’t in the database.

If we let the AI curate our entire reality, we lose the serendipity. We lose the thrill of the hunt. We become tourists in our own lives, guided by a machine that thinks the best restaurant is the one with the most stars.

We have to build systems that allow us to get lost again.

Status: > RUNNING REALITY_CHECK.EXE Result: ERROR: AUTHENTICITY NOT FOUND IN TRAINING DATA

The web might be dead, but we aren’t. Be the glitch in the algorithm. Make them cite you.