At first, it is prudence. Then it becomes strength. Then, quietly, it begins to look like a contraction disguised as efficiency.
We are somewhere inside that transition now.
Block announced a roughly 40% workforce reduction as part of restructuring tied to efficiency and AI leverage. Amazon has continued rolling layoffs after cutting more than 27,000 roles in recent cycles. Meta eliminated more than 20,000 during its “Year of Efficiency.” Alphabet cut 12,000. Microsoft trimmed 10,000+. These were not cosmetic adjustments. They were structural resets.
Markets rewarded them. Margins improved. Earnings stabilized. Stocks re-rated.
Inside a narrow lens, the math works; labor is often the largest operating expense. Reduce it, expand margins, improve forward earnings.
But that lens ignores the system those margins exist inside.
The real differentiator isn’t who understands AI. It’s what percentage of your workforce can wield it well enough to eliminate dependency.
Every leadership team now claims to “get” AI. That is no longer differentiating. The real dividing line is distribution.
What percentage of your workforce can use AI not just as assistance, but as force multiplication?
There has always been a difference between those who tread water and those who pull the ship. AI doesn’t create that difference. It exposes and amplifies it.
When one operator can now produce what previously required three, the internal question shifts from “Who do we need?” to “How few do we truly require?”
Markets love that question in the early phase.
But there is a second dynamic unfolding at the same time.
The Red Queen Problem
If everyone doubles their output, doubling output becomes the baseline. This is the Red Queen problem: you must run faster simply to stay in place.1
AI does not “save time” in competitive markets. It raises the survival threshold. What once differentiated you becomes table stakes. What once justified your headcount becomes excess.
In that environment, productivity gains do not necessarily translate into leisure, or slack, or even stability. They translate into escalation.
One firm uses AI to reduce costs and improve margins. Competitors must follow or lose pricing power. Soon, the entire sector operates at a higher output-per-employee ratio, and the previous staffing model looks bloated everywhere.
Efficiency becomes mandatory.
And mandatory efficiency eventually collides with demand.
The Consumption Concentration Problem
According to Moody’s Analytics and Federal Reserve data, the top 10% of U.S. earners account for roughly 45–50% of total consumer spending2. That concentration has increased over decades as income distribution skewed upward.
Now consider who is disproportionately represented in white-collar layoffs:
Engineers.
Senior managers.
Finance professionals.
High-salary knowledge workers.
These are precisely the households that drive discretionary consumption: travel, technology upgrades, restaurants, services, and housing renovations.
When a $350,000 household transitions into severance or unemployment, the adjustment is not theatrical. It is incremental. Vacations are delayed. Large purchases are deferred. Risk appetite declines.
One household does not move GDP. Tens of thousands begin to.
And this is where the market narrative grows incomplete.
The Elasticity Question
The first-order effect of layoffs is obvious: Lower expenses → improved margins → stronger earnings per share.
Markets understand that.
The second-order effect is slower: High-income layoffs → reduced discretionary spending → revenue softness → broader earnings pressure.
The market has priced the discipline. It has not yet priced the elasticity.
The missing variable is elasticity, how sensitive aggregate demand is to income compression among the top decile of earners. When nearly half of consumption sits inside the top 10%, demand elasticity is not theoretical. It is material.
Markets have priced the cost savings. They have not fully priced what happens if high-earning households meaningfully retrench.
That is the line we are walking.
Historical Echoes
After the dot-com collapse in 2001, the NASDAQ fell nearly 78% peak to trough. Tech employment contracted sharply. Re-employment happened, but compensation compressed and the sector emerged leaner relative to GDP.
After 2008, the financial sector shrank materially. Many professionals were reabsorbed, but the industry’s footprint and compensation growth did not return to the pre-crisis trajectory for years.
Re-employment occurred, but the total addressable market for certain high-earning roles narrowed. The question for 2026 is whether AI is a tool for those roles or a replacement for the headcount itself. That distinction matters today.
If AI permanently reduces the number of mid- and senior-level knowledge roles required per unit of output, we are not witnessing cyclical layoffs. We are witnessing structural resizing.
The macro outcome hinges not just on how quickly displaced workers find new jobs, but whether those jobs exist at similar compensation levels.
Are they migrating into new sectors that expand total output?
Or is the TAM for high-earning white-collar roles compressing?
The Bull Case Deserves Respect
There is a legitimate expansion thesis.
Historically, when the cost of production falls, demand often rises. Jevons Paradox teaches that efficiency can increase total consumption of a resource, not decrease it.3 Cheap computing did not shrink computing. It created the internet economy.
The bullish case says AI will unlock new industries, new markets, and new revenue pools. Firms will not cut two workers and coast. They will keep three and produce nine times as much. Output expands. GDP rises. Displaced workers redeploy into higher-order functions.
That future is possible.
But it depends on elasticity being expansionary. If AI-driven efficiency fuels new demand, the system stabilizes.
If it primarily fuels cost competition and margin defense, the gains concentrate rather than expand.
Where Is the Equilibrium?
Too little cutting and margins suffer.
Too much cutting and demand weakens.
The equilibrium is narrower than it appears.
AI is accelerating internal sorting inside companies. The next phase will sort the broader economy. Talent density, demand elasticity, and the velocity of re-employment will determine whether this period becomes a productivity renaissance or a consumption compression cycle.
Markets are currently rewarding discipline.
They have not yet fully priced elasticity.
And if the Red Queen dynamic turns efficiency into a survival baseline rather than a growth catalyst, the line between optimization and contraction will get thinner with each earnings season.
We are not debating whether AI changes productivity. We are debating whether it changes demand.
That is the macro question underneath the margin story.
Wondering how to position yourself in a market like this? Read our next article, where we outline our approach and provide a roadmap.