Written by Asi de Silva CFA | March 11, 2026
Geopolitical shocks such as the recent Iran crisis are disruptive because they are immediate, visible, and emotionally gripping. But the risks that matter most for long-term investors are not always the ones dominating headlines. As markets calm, I believe it is worth turning to a more structural question: whether AI, for all its firm-level productivity benefits, could create second-order pressures on labor income, demand, and ultimately market valuations.
One recent report helped crystallize this debate for many investors.
Citrini Research’s recent note titled The 2028 Global Intelligence Crisis can be best described as a thought experiment on the possible negative economic outcomes from rapid AI adoption.
The report quickly went viral. The Wall Street Journal’s coverage, Viral Doomsday Report Lays Bare Wall Street’s Deep Anxiety About AI Future, reflects how deeply this question is resonating across institutional circles.
Just days later, Block Inc. announced a 40 percent workforce reduction. I view Jack Dorsey, Block’s founder and CEO, as a long-term thinker and technology visionary.
His decision does not validate Citrini’s conclusions outright, but it does make the underlying debate more tangible. When forward-looking operators begin aggressively restructuring around AI-driven efficiency, the reflexive implications move from theory toward practice.
That said, at CSF, we have a more sanguine view on the likely path for AI and economic activity.
Our base case is that widespread AI deployment will encounter meaningful physical and industrial bottlenecks, including power generation, grid capacity, semiconductor fabrication, and broader infrastructure constraints. Those limitations could temper the speed and breadth of the displacement dynamic that Citrini’s more extreme scenario assumes.
We also recognize that this technological shift is different from prior cycles. The disintermediation of intelligence is not the same as the automation of manufacturing or the digitization of services. We are navigating unfamiliar waters, and historical analogues may only partially guide us.
In short, we are taking the debate seriously without adopting the most extreme conclusions.
What Exactly Did The Citrini Report Say?
Here’s how Google Gemini AI bot distilled the article for a “retail investing audience” (based on my prompt):
“we are entering a "reflexive loop" where AI’s success as a tool for corporate efficiency becomes a failure for the broader economy. Citrini argues that as companies use "Agentic AI" to slash white-collar headcounts and boost short-term margins, they inadvertently destroy the consumer income base (the Intelligence Displacement Spiral) that fuels their own long-term revenue. This creates "Ghost GDP"—high productivity on paper that doesn't circulate in the real world—eventually leading to a systemic collapse of discretionary spending, a crisis in the mortgage market, and a projected 38% to 57% drawdown in the S&P 500 by 2028.”
In other words, the core thesis is not that AI will fail. It is that AI may succeed at the firm level in ways that destabilize the broader economy.
Citrini is not bearish AI. It is Macro Skepticism.
Their argument is not technological skepticism. It is macro skepticism.
Specifically, what happens if AI functions as a General Intelligence Substitute rather than a complement?
If there is no safe harbor for displaced cognitive labor and no new sector absorbing white-collar workers, productivity gains may not recycle into income growth.
That is the reflexive loop at the heart of their scenario.
Micro (company) Rationality Leading to Macro (overall economy) Strain
From the perspective of an individual firm, the incentives are clear:
Replace analysts with AI agents.
Replace customer service teams with automated workflows.
Flatten middle management.
Expand margins.
That is rational corporate behavior.
But if enough firms pursue this path simultaneously, the macro picture changes:
Aggregate white-collar wages decline.
Discretionary spending slows.
Mortgage formation weakens.
Credit velocity falls.
Citrini describes this divergence as Ghost GDP. Output statistics remain elevated while money circulation and economic activity deteriorate.
Block’s 40% Workforce Reduction Adds Fuel to The Debate
In the same week, Jack Dorsey, the founder of Twitter & CEO of Block, announced roughly 40% workforce reductions at Block Inc, the parent of Square & Cash App.
Dorsey framed it as a strategic response to how AI changes what smaller teams can accomplish.
Here’s a paragraph from Jack’s letter to employees pertinent to our discussion:
You can read the full letter on Jack’s X account
Importantly, Block was not cutting from a position of obvious crisis, which is part of why the announcement carried signal value in markets. The analysis below illustrates the point.
Other opinions suggests that Block was bloated and that Dorsey, while a visionary, is a poor manager. Both views are valid, judging by Block’s weak share price since 2021.
However, since the staff reduction announcement, the stock is up nearly 30%. This type of reaction may temp other CEO to try to engineer similar outcomes/reactions. A one week share price move doesn’t have much signal, but we’ll be watch this space closely.
Source: Via X
Let’s take Dorsey at face value that the action is strategic rather than reactive.
If AI materially increases productivity per employee, forward-looking CEOs may reduce labor to boost margins. After all the majority of CEO compensation is stock options.
That is precisely the kind of micro-rational behavior Citrini is modeling.
One company doing this improves margins.
Many companies doing this could compress aggregate income.
Why We See a Lower Probability of The Citrini Scenario
My more sanguine outlook is underpinned by AI voracious appetite for resources. The world is likely to encounter physical bottlenecks as we are already witnessing:
Power generation capacity
Grid infrastructure
Semiconductor fabrication limits
Industrial supply chain constraints
AI is not only software. It requires compute, electricity, copper, transformers, and turbines.
On the supply side, price signals for higher resource output take time to translate into capacity, often measured in years and sometimes decades. There is little reason to assume this cycle will be frictionless.
This is one reason I continue to hold meaningful exposure to commodities and commodity-linked equities, a positioning that is partially underwritten by AI-driven infrastructure demand.
Ultimately, if the transition of white-collar work to AI unfolds gradually over a longer period, the odds improve that the global economy can adapt with less disruption. If, on the other hand, the substitution is rapid and concentrated into a short window, as Citrini’s scenario envisions, we should expect more serious dislocations.
Why This Matters for Investors
Today’s equity market assumes in aggregate:
AI drives margin expansion.
AI drives revenue acceleration.
AI drives durable earnings growth.
This is one reasons valuations of US stock indices remain elevated relative to history.
Citrini asks whether margin expansion may undermine aggregate revenue growth. If aggregate income contracts faster than productivity expands, revenue growth may disappoint even as margins initially improve.
That produces a very different earnings trajectory than current multiples imply.
We have already seen the market starting to distinguish between AI-exposed sectors and those that a less exposed.
For example, the US software sector ETF (IGV) is down ~ 21% in 2026.
Goldman has coined HALO = Heavy Assets and Low Obsolescence for stocks with low AI risk.
It’s too early to know if software revenues are structurally impaired, but that was an expensive sector we had already avoided.
What Does it Mean For You?
This is not a prediction of collapse.
It is a reminder that technological revolutions often create second-order effects markets initially overlook.
The internet boom did not fail because the technology did not work. It failed because expectations (and valuations) exceeded the economy’s capacity to absorb it.
Citrini is asking whether AI may follow a similar reflexive arc, not because it disappoints technologically, but because it disrupts income formation faster than the system can adapt.
For our clients, we will be reading, researching, analyzing and adapting portfolios as the outlook & investment implications become clearer.
Final Thoughts
I do not read Citrini as anti-AI.
I read them as anti-extrapolation.
When a visionary CEO like Jack Dorsey makes aggressive labor cuts in anticipation of AI-driven productivity, it suggests this debate is no longer theoretical.
I strongly recommend reading the full note:
https://www.citriniresearch.com/p/2028gic
Estimated read time:
10 to 15 minutes for seasoned macro and technology investors
20 to 25 minutes for readers less familiar financial markets, labor market reflexivity or mortgage market dynamics
It is provocative.
It is uncomfortable.
And it is worth thinking about carefully.