In February 2026, a Substack essay written as a fictional macro memo from June 2028 triggered real market selloffs across software, legal services, cybersecurity, wealth management, and insurance sectors. The piece, published by Citrini Research, modeled what happens when AI capability compounds faster than institutions can adapt. It described a five-stage displacement spiral ending in a repricing of the assumption that has underwritten every institution in the modern economy: that human intelligence is scarce.
The essay was speculative. The market reaction was not.
That asymmetry is worth examining, because it reveals something about institutional behavior that is more consequential than any individual prediction about AI: institutions do not respond to what is happening. They respond to the narratives they consume about what is happening. And when the narrative is misdiagnosed, the institutional response accelerates the very crisis it fears.
The Citrini essay is brilliantly constructed. It is also fundamentally misdiagnosed.
"Institutions do not respond to what is happening. They respond to the narratives they consume about what is happening."
The Five Stages, Reframed
The displacement spiral Citrini describes unfolds in five stages. Each one is real. Each one is also misnamed.
Stage 1: The Capability Leap
Citrini describes agentic coding tools enabling in-house replication of mid-market SaaS products, collapsing software differentiation and triggering procurement renegotiations. This is accurate. What is missing is the evaluative question: differentiation in what?
Software companies that compete on features will face this compression. Features are knowing: discrete capabilities that can be listed, compared, and replicated. But institutions do not fail because they chose the wrong features. They fail because they never measured whether the technology produced institutional outcomes across financial sustainability, achievement impact, equity exposure, and operational durability. The capability leap does not eliminate software. It eliminates the ability to hide behind feature lists instead of institutional evidence.
Stage 2: The Reflexive Loop
Companies cut headcount, reinvest savings in AI, which enables further cuts. Individual rationality produces collective catastrophe. Citrini calls this a feedback loop with no natural brake.
The loop is real, but the diagnosis is incomplete. The reflexive loop is not a technology problem. It is an institutional integrity problem. Organizations trapped in the loop are defending territory: protecting margins, headcount ratios, and quarterly metrics rather than asking what the institution actually produces that cannot be automated. The organizations most vulnerable to the reflexive loop are the ones that cannot distinguish between institutional outputs (reports, tickets, deliverables) and institutional outcomes (thinking, agency, adaptive capacity). The loop has a brake. The brake is institutional self-knowledge. Most organizations do not have it.
Stage 3: Friction Goes to Zero
AI agents eliminate intermediation. Business models built on human friction — subscriptions, inertia, habitual loyalty — disintegrate. Interchange fees, commissions, and service premiums compress.
This is the stage that should unsettle institutions the most, because it asks a question most organizations have never confronted: which of our structures depend on friction, and which produce genuine relational value? An institution where authority and accountability are misaligned is one whose value proposition is friction with a friendly face. When friction disappears, those institutions have nothing underneath. The ones that survive are the ones where authority, accountability, communication, belonging, and the way the organization relates to time are structurally aligned — not just culturally performed.
Stage 4: Ghost GDP
Output shows up in national accounts but never circulates through the real economy. Productivity surges while consumption withers. Citrini calls it Ghost GDP: the human-centric economy decouples from the machine-centric economy.
Education has its own version of this, and it predates AI by decades. Call it Ghost Learning: completion rates and test scores that appear in accountability metrics but never translate into thinking, agency, or institutional capacity. Standardized assessments measure recall and compliance. They do not measure whether a student can hold contradiction, generate meaning from complexity, or exercise judgment under uncertainty. The displacement spiral does not create Ghost GDP. It reveals that we have been producing ghost outcomes across every institution for years and calling them real. AI is just the reagent that makes the gap visible.
Stage 5: The Intelligence Premium Unwind
The assumption that human intelligence is scarce collapses. Every institution built on that assumption reprices. The financial system, labor market, mortgage market, and tax code all face structural failure.
This is the essay's most important stage and its deepest misdiagnosis.
The intelligence premium is not unwinding because AI replicates human intelligence. It is unwinding because our institutions confused knowing with thinking. Hannah Arendt drew this distinction in 1978: knowing is the accumulation and application of established information. Thinking is the capacity to examine, question, and generate meaning from what is not yet known. They are fundamentally different cognitive activities. An institution can be full of knowledgeable people and produce no thinking at all.
AI replaces knowing. It cannot replace thinking. The institutions that produce thinking — genuine inquiry, structural self-examination, the capacity to hold complexity without collapsing into compliance — are resilient to the displacement spiral. The institutions that warehouse knowing are exposed.
"The intelligence premium was never about intelligence. It was about the institutional conditions that allow thinking to occur."
What the Market Reaction Reveals
A fictional essay moved real markets. That fact deserves as much analysis as the essay itself.
The market reaction reveals that institutional decision-makers recognized the displacement spiral as plausible because they already sense it in their own organizations. They feel the reflexive loop forming. They see the friction that props up their value propositions. They know their metrics are closer to ghost outcomes than they want to admit.
But the response was financial — sell, hedge, reprice — rather than structural: examine, redesign, rebuild. This is the pattern. Institutions respond to disruption narratives by protecting positions rather than examining the conditions that make them vulnerable. The Citrini essay describes a crisis that no existing framework can address. Its closing line is an admission: "Nobody's framework fits, because none were designed for a world where the scarce input became abundant. So we have to make new frameworks. Whether we build them in time is the only question that matters."
That closing line is more honest than most institutional leaders are willing to be. It is also incomplete.
Frameworks for evaluating institutional integrity under structural disruption do exist. They measure whether an organization produces thinking or warehouses knowing. They diagnose whether institutional structures depend on friction or produce genuine value. They assess financial sustainability, achievement impact, equity exposure, and operational durability simultaneously, because single-dimension evaluation is how institutions hide from the questions that matter.
"The financial world has named the disease. The question is whether anyone is willing to use the treatment protocol."
The Education Sector's Unique Position
One sector is conspicuously absent from the Citrini analysis: education. This is not surprising. Education is routinely excluded from cross-sector economic modeling, partly because external researchers overlook it and partly because the profession has long resisted being measured on the same terms as everyone else.
But education is where the displacement spiral either breaks or accelerates. If schools and universities continue producing knowledge workers who merely know, they are producing the very workforce AI will replace. If they produce leaders who think, they are building the only human capacity that remains scarce.
The sector most responsible for preparing every other workforce for AI is not in the conversation. That is not a gap in the analysis. It is the gap that makes the analysis incomplete.