The Four Species of Web Content

When we started LifeByLogic in late 2025, we had a single uncomfortable question: what should a publication even look like when ChatGPT summarizes the entire web in three seconds? The 1,200-word explainer — the default shape of a web page for two decades — suddenly felt obsolete. Google's AI Overviews were launching. Gartner had predicted a 25% collapse in search traffic by 2026. Publishers were reporting 40 to 70 percent losses in organic traffic, year over year.

Every answer to "how should we publish" assumed the article was still the unit. We came to believe it was the wrong unit. So we spent three months building a framework before we built a site. What emerged shaped every design choice behind LifeByLogic — from the four labs, to the tool-first structure, to the explicit welcome we give to every AI crawler.

We call it the Species Taxonomy. It is our thesis and our operating manual, published here so it can be used, argued with, or discarded by anyone else facing the same question.

§I. What is the Species Taxonomy?

The Species Taxonomy is a framework for classifying every piece of web content into one of four categories, based on two axes. The horizontal axis measures AI substitution risk — how easily a language model can replicate the content. The vertical axis measures originality of value — whether the content creates new information or repackages what already exists.

Plot those axes against each other and four quadrants appear. Each quadrant holds one species of web content. Each species has a fate. Two are being replaced by AI. Two are becoming the only things worth making.

§II. Why are articles dying in 2026?

Species I content is dying because AI now answers derivative queries directly in the search interface, eliminating the click. The data is no longer subtle. According to Bain & Company, 60% of US Google searches ended without a click in early 2025. According to Semrush, searches that trigger Google's AI Overviews now show an 83% zero-click rate, and searches in Google's new AI Mode reach 93%.

This is the actuarial table for Species I. If you operate a site whose content is mostly "What is compound interest" or "How to improve your sleep," your content is now what Pew Research called "the 1% problem": only 1% of users click any link inside a Google AI Overview.

Gartner's 2024 prediction of a 25% decline in search traffic by 2026 now looks conservative. Individual publishers have reported 40–70% organic losses in a single year. News sites — which are the heaviest producers of Species I — have been hit hardest.

When we planned LifeByLogic, this was the central fact we refused to dismiss. Every publisher we talked to was treating the shift as a traffic problem. It is not. It is a structural problem. Species I competes with AI. You cannot out-publish an engine that writes 10,000 explainers a second.

This is why LifeByLogic publishes no "Top 10 Sleep Tips," no "What is REM sleep," no "How to Improve Focus." We chose not to write any of them. Those articles are not bad; they are the wrong category of thing to make in 2026. Instead we built the Sleep-Cognition Optimizer, which computes your sleep schedule from your chronotype. That is Species III, and nothing in AI's architecture can replace it.

§III. Why do the other three species survive?

Species II, III, and IV share one property: AI either needs them, is amplified by them, or cannot produce them. The reasons differ; the consequence is the same. Traffic shrinks for Species I while AI systems — the same systems replacing Species I — increasingly point people toward Species II, III, and IV.

Species II — Primary Knowledge survives because AI needs source material.

AI is, mechanically, a compression algorithm for human-produced information. Compression requires uncompressed input. Every AI summary of a clinical trial traces back to a research team that spent three years on it. Remove the primary layer and AI collapses into self-quotation, inventing studies that do not exist.

LifeByLogic produces Species II wherever it can. The methodology behind the Brain Age Index cites the 2024 Lancet Commission on Dementia Prevention and twelve other primary sources. Our upcoming essay, "The 17 Factors That Age Your Brain," synthesizes primary literature into a framework we hope becomes reference material. This is why we publish cited methodology: Species II content signals to AI that we should be cited.

Species III — Interactive Instruments thrive because AI cannot be an instrument.

An Interactive Instrument is a web content format that computes a personalized result from user inputs. Calculators, simulators, assessment tools, configurators. AI can describe what a mortgage calculator does. AI cannot be a mortgage calculator.

The distinction matters because the user wants the number, computed from their specific inputs — not a description of how the number would be computed. When someone asks for their cognitive age, they do not want an article about cognitive aging. They want to enter their factors into the LBL-BAI and see their personal result.

A calculator is not a summary. A simulator is not a paragraph. The value is the computation, not the prose around it. This is why Brain Lab and Crossroads Lab are built around instruments with named methodologies. It is also why the All Tools page lists 21 instruments, not 21 articles.

Species IV — Identity & Community thrive because AI has no identity.

An AI response is, by design, a compressed average of a million voices. It has no point of view. The moment you ask an AI for its opinion — "what is your favorite novel?" — you get a committee answer no human would give. This is not a flaw. It is the format.

The inversion: identity, which was always the most valuable thing a publication could have, becomes the only thing AI cannot produce. LifeByLogic leans into this deliberately. Our editorial anonymity, our six commitments, the insistence on named methodology (LBL-BAI, LBL-SCO, LBL-QS, LBL-BD, LBL-LS) — all Species IV. The publication has an identity precisely because AI cannot.

§IV. What does the new hierarchy look like?

The old web put articles at the top of the hierarchy. The new web puts instruments and identity on top, with derivative content as noise.

On the old web, articles drove traffic, traffic drove ad revenue, ad revenue funded more articles. Tools and databases were footnotes — useful, but not the product. On the new web, tools are the product. Articles are the connective tissue around them.

This has consequences for every publication. A news site producing only Species I summaries of events will not survive the decade. A site producing Species II — original reporting — becomes more valuable as AI summarizes its work into broader distribution. A site producing Species III instruments captures users who skip the summaries entirely. A site built on Species IV — voice, trust, subscription — is insulated from AI because it sells what AI structurally cannot.

The worst possible strategy, and the most common, is to double down on Species I because the business already knows how to make it. This is where the 40–70% traffic losses come from. The content isn't worse than it was. It has merely moved into the quadrant AI handles.

§V. What we built at LifeByLogic.

LifeByLogic is a deliberate experiment in building a publication entirely around Species III and IV. Every piece on the site fits one of the four quadrants, by design. Here is the mapping:

Species III — the two live tools so far: The Brain Age Index (LBL-BAI), a 17-factor calculator grounded in the 2024 Lancet Commission. The Sleep-Cognition Optimizer (LBL-SCO), which assigns your chronotype and generates a personalized sleep schedule. In development: the Crossroads Lab's Should I Quit? simulator, the Behavior Lab's Cognitive Bias Detector, and the Life Dashboard's composite Life Score. All Species III. All with named methodology, cited research, and computation that AI cannot replicate.

Species IV — the editorial architecture: The anonymous editorial team, the institutional voice, the six commitments, the blog where you are reading this essay, the upcoming Logic Letter newsletter. A publication with a clear identity, exactly because AI has none.

Species II — the methodology and this essay: Every tool publishes its full methodology, cited to primary sources. This essay is Species II (it introduces a framework that did not previously exist) wrapped in Species IV (the institutional voice).

Species I — none: We publish no derivative content. No listicles. No explainers. Not because we look down on them, but because we believe they are the wrong thing to make in 2026.

We also welcome every AI crawler explicitly — our robots.txt lists GPTBot, ClaudeBot, PerplexityBot, Google-Extended, CCBot, and more — because we are not competing with them. They link to us because we build what they cannot. Be the thing AI references, not the thing AI replaces.

Whether you operate a publication, a business, a research lab, or a personal blog, the question the Species Taxonomy forces you to answer is the same. Which species are you producing? Which quadrant does your work live in? If most of your output is Species I, the 93% number is your future. If it is Species II, III, or IV, the decade ahead is the most interesting the web has ever had.

The last article in the Species I sense has already been written. It just hasn't been published yet.