What Happens When Productivity Soars but Output Doesn’t?

Exploring the Cobb-Douglas Equation in an Age of AI Disruption

As AI accelerates productivity, our capacity to generate value—code, research, and decisions may outpace what the economy is actually ready to absorb. Economists have a framework for thinking about this: the Cobb-Douglas production function.

Y=A⋅K^α⋅L^1−α

Where:

• Y is total output
• K is capital (machines, tools, infrastructure)
• L is labor (workers, hours)
• A is total factor productivity (TFP), the secret sauce of how efficiently capital and labor turn into value
• α is the output elasticity of capital (typically between 0.3 and 0.4)

It tells us something deceptively simple: output is a function of inputs (capital and labor), and how efficiently you use them (productivity).

We’ve lived in a world where capital and labor were the center of economic effort. But what happens when A(productivity) suddenly takes a quantum leap forward, thanks to AI?

When A Goes Up, What Should Happen?

In theory, if productivity (A) rises, output (Y) should rise too, even with the same K and L. That’s what innovation promises: more from less.

But what if A surges, and Y doesn’t?

Maybe demand doesn’t increase. Maybe AI efficiency is captured by a few firms. Maybe regulation or inequality dampens the spread.

In Cobb-Douglas math, if A increases but Y stays flat, then either K or L (or both) must fall to keep the equation balanced. The economy now needs less labor or less capital to produce the same output.

This is where things get strange—and deeply relevant.

Flashback to the Pandemic: Output Fell, but Productivity Didn’t

In early 2020, U.S. output plummeted. GDP dropped nearly 33% on an annualized basis in Q2. But that wasn’t because productivity collapsed. It was because the machine shut down—factories idled, services paused, and people stayed home.

• Capital utilization fell to record lows (~64%)
• Labor participation dropped to 60.2%, the lowest in 50 years

The economy had the tools and knowledge to produce, but it couldn’t. This was a utilization shock, not a productivity shock. Although the nature of the shock is very different from what an AI-driven productivity surge might look like, we can still learn from how capital (K) and labor (L) adjusted to absorb the disruption. Businesses rapidly paused or repurposed capital, while labor reallocated—sometimes temporarily, sometimes structurally. As we now face the prospect of a productivity boom outpacing demand, similar adjustments may be underway: capital being sidelined or redirected, and labor seeking alternative paths to stay productive or solvent in the short term.

Inverse Shock Ahead: AI Boosts A — But Demand Trails

Imagine a future where AI significantly increases productivity, but demand and income distribution do not rise in sync. In that case:

• Capital (K) becomes underutilized. Why build new factories if you can produce more with less?
• Labor (L) gets displaced without enough new jobs to absorb workers.
• Prices come under deflationary pressure due to excess supply.
  

This “inverse disruption” replaces pandemic-style shortages with a glut of unsold capacity—machines idle, people waiting for roles that no longer exist.

Redeploying Idle Capital—and Why It’s Harder This Time

In a future AI-driven surge, we face the inverse: excess capacity, but no obvious crisis to mobilize it against. Worse, today’s capital markets already show signs of strain:

• While big tech firms like Microsoft, Amazon, and Alphabet are doubling down on AI infrastructure, broader corporate capex is leveling off about 12–14% of GDP, flat over the past year.
• According to Deloitte, small-business capital spending dipped to its lowest since 2020 in April and remained soggy in May.
• Larry Fink of BlackRock noted that $25 trillion in idle cash sits in banks and money market funds, the highest in his career.
  

The $25 Trillion Question: Why Capital Sits Idle During an AI Revolution

Larry Fink’s observation about $25 trillion in idle cash, nearly equal to the entire U.S. GDP, represents a massive paradox. That’s 250x annual global VC deployment, yet it sits in money markets earning 5% while AI transformation begs for investment.

The Paralysis Problem

Risk-Return Mismatch: Why chase speculative 10% returns when you get 5% risk-free? This calculation freezes capital exactly when we need it mobilized.

AI Investment Paradox: While Big Tech pours billions into AI, most corporations watch from the sidelines, paralyzed by:

• Unclear ROI (unlike cloud computing’s obvious returns)
• Fear of winner-take-all dynamics
• Analysis paralysis (“Wait for GPT-6?”)

The Innovation Bottleneck

Traditional VC is breaking under AI’s weight. The old model where $10M got you to product-market fit is dead—that barely covers GPU costs now. This creates a “missing middle” where startups can’t access the $50-100M needed to compete without crushing dilution.

Meanwhile, Fortune 500 companies sit on record cash, unable to decide to build (massive costs), buy (sky-high valuations), or wait (risk obsolescence).

Unlocking the Trillion-Dollar Vault

We need financial innovation to match our technical innovation:

• New instruments: AI infrastructure bonds, productivity-linked securities, tokenized compute
• Public-private partnerships: Even 1% of idle cash could fund national AI labs, open infrastructure, and retraining programs
• The SpaceX model for AI: Breakthrough business models that democratize access and create clear monetization paths

Every day this capital sits idle represents lost productivity gains and widening competitive gaps. The tragedy isn’t lack of money—it’s lack of mechanisms to deploy it during history’s most transformative moment.

The bottom line: We have the cash. We have the technology. We lack a bridge between them.

Labor and a Surprising Lifeline: Retail Trading

When the pandemic hit, and millions of workers were suddenly furloughed or laid off, a surprising behavioral shift occurred: retail trading surged. With job markets frozen and many traditional avenues of income disrupted, people turned to the stock market in droves—some out of financial necessity, others out of curiosity, and many simply because they had the time and tools to do so. Platforms like Robinhood, Webull, and SoFi saw record sign-ups, fueled by a perfect storm of stimulus checks, zero-commission trading, and viral finance culture on Reddit and TikTok.

In fact, between March and June 2020, Robinhood reportedly added over 3 million new users. At the same time, retail investors were estimated to account for up to 25% of all equity trading volume on certain high-volatility days. Trading became a social phenomenon—GameStop and AMC weren’t just tickers, they were cultural moments.

Apps like Robinhood exploded. Millions entered markets, some for the first time. For many, trading became:

• A side hustle
• A coping mechanism
• A way to feel agency

But it was risky. Some made money. Many lost it. When the market cooled, a lot of those gains vanished. Inexperienced traders took huge losses.

So, should we dismiss retail trading? Not entirely.

The Data: A Boom & Its Echoes

• U.S. equities volumes jumped ~62% to ~11 billion shares per day post-pandemic—retail played a huge role, especially in low-priced stocks, where they represent ~9% of trade volume and ~6.5% by dollar value, above pre-COVID norms (ref: Growth of US equities Volumes and Rise of Retail)
• Retail activity surged dramatically during COVID; options trading participation hit a pandemic high of ~48%, later stabilizing in the mid-40% .
• In H1 2025, investors, many retail, traded a record $6.6 trillion in U.S. equities, with nearly $155 billion net inflows, surpassing 2021 meme-stock levels.

The Opportunity

• Access to markets democratized financial participation
• Potential for earnings, especially amid volatility
• Engagement and agency for displaced workers, giving purpose during downtime
  

The Risk

• High losses from inexperience or compounding financial stress
• Too much speculation (e.g., surge in penny stocks over 47% of daily volume by June 2025
• Emotional strain from volatile returns and FOMO
• Exposure to scams or fraud, amplified in retail-heavy environments
  

A Better Playbook

Building a safer bridge means:

• Educational tools (probability-based trading training, risk management)
• Data-driven analytics and clear risk visualizations
• Regulatory guardrails for margin, disclaimers, and trade limits
• Supportive communities to curb impulsive behavior and promote best practices
  

When done right, trading becomes a skill-building activity, not a high-stakes gamble.

Final Words: From Abundance to Activation

AI might gift us a world of astonishing productivity, but that alone doesn’t guarantee progress. If history has taught us anything, it’s that disruptions, whether sudden like a pandemic or gradual like automation, demand foresight, adaptability, and bold public imagination.

The real challenge isn’t whether we can produce more. It’s whether we’ve learned enough from past shocks to channel the surplus wisely. That means preparing for mismatches between what we can supply and what society can absorb before they happen.

Will we build the institutional muscle to redeploy idle capital? Will we equip workers not just with skills, but with platforms and protections to navigate the transition? Will we invest in ideas and infrastructures that transform abundance into shared resilience?

We’re not at the end of the road, we’re at a fork. And the sooner we embrace this as a planning problem, not just a production puzzle, the more likely we are to turn productivity into prosperity.