When AI Success Becomes Economic Suicide

A blog post just vaporized $200 billion in stock value.

Last Sunday, a little-known firm called Citrini Research published what looked like science fiction: a memo "from June 2028" describing how AI triggered a global economic meltdown. By Monday, DoorDash, Uber, and other companies mentioned in the post saw their shares tumble as investors panicked.

The twist? In Citrini's scenario, AI doesn't crash the economy by failing — it destroys it by succeeding too well.

The Productivity Paradox

Forget everything you've heard about AI being overhyped. Citrini's nightmare begins with AI delivering exactly what its boosters promised: productivity growth unseen since the 1950s, mind-boggling profits, and massive GDP gains.

But here's where things get dark.

As AI agents become capable enough to replace a $180,000 product manager for $200 per month, companies across consulting, software, real estate, financial services, and legal work start slashing headcount. The savings? They plow it all back into more AI.

This creates what economists call a doom loop. More AI investment leads to more capable agents. More capable agents displace more white-collar workers. Displaced workers slash their spending and flood into blue-collar jobs, driving down wages across the board. Meanwhile, the massive profits from AI productivity gains flow to an increasingly narrow elite.

Sam Altman can only buy so many cars and TVs. When the ultra-wealthy get richer, much of that money doesn't circulate back through the economy. Consumer demand collapses just as productive capacity soars.

Companies respond to falling demand by cutting costs — which means more AI investment and fewer human workers. The cycle accelerates with no natural brake.

The Great Rent Collapse

Citrini's second story focuses on how AI agents will obliterate entire business models by perfecting comparison shopping.

Today's economy runs on human laziness. We don't have time to research every purchase exhaustively, so we default to familiar brands or whatever appears first in search results. This "search friction" has allowed companies to extract trillions of dollars in economic rents.

AI agents don't get impatient. They can instantly compare prices across the entire internet and build competing services overnight.

Imagine this scenario: Bob, armed with Claude Code, builds a delivery platform in an afternoon that undercuts DoorDash on fees. In today's world, Bob's app would struggle to gain traction — network effects protect incumbents.

But when everyone uses AI agents that automatically find the cheapest option, Bob's platform can replicate DoorDash's network in days. Restaurants, drivers, and customers all get routed to whoever offers the best deal.

The same dynamic plays out across insurance (exhaustive comparison shopping becomes effortless), enterprise software (companies can build custom solutions or choose from countless AI-generated alternatives), and real estate (traditional brokerages become obsolete as AI eliminates information asymmetries).

Margins collapse. Companies accelerate the "layoffs → AI investment → lower demand → more layoffs" cycle.

Why Markets Freaked Out

It's remarkable that a speculative blog post could move markets so dramatically. Several factors explain the reaction:

First, the scenario felt uncomfortably plausible to many investors already nervous about AI's disruptive potential. Unlike typical doom-and-gloom predictions, Citrini's memo included specific companies, timeframes, and mechanisms.

Second, the memo's counterintuitive logic — that AI success could trigger economic collapse — caught many off guard. Most investors have been focused on whether AI companies can justify their valuations, not whether their success might destabilize the entire economy.

Third, the current moment feels genuinely uncertain. ChatGPT went from zero to ubiquitous in two years. AI agents are already handling real business tasks. The pace of change makes even seasoned traders struggle to distinguish science fiction from emerging reality.

The Scenario's Weak Points

But there are compelling reasons to doubt Citrini's narrative, at least in its full form.

AI may not cause mass unemployment. Despite years of generative AI adoption, US unemployment remains near historic lows. Even AI-exposed professions like software development have seen job openings increase. Every previous general-purpose technology eliminated some jobs while creating others.

Investment money doesn't vanish. When AI companies spend $200 billion quarterly on infrastructure, that money flows to construction workers, electricians, engineers, and lawyers who then spend it in their local economies. Citrini simply asserts this circulation doesn't happen without explaining why.

Collapsing rents would boost demand. If AI forces down prices across industries, that effectively redistributes income from business owners to consumers. Working-class Americans spend a higher share of their income than wealthy shareholders, so this redistribution would increase total consumer spending.

Governments would likely intervene. When millions of politically influential professionals lose their jobs simultaneously while productive capacity soars, politicians would face enormous pressure to redirect AI profits through taxation and redistribution.

Building trust takes time. Sure, Bob can code "DoorSprint" overnight, but providing reliable customer service, logistics optimization, insurance, and fraud protection isn't trivial. And when AI can mint scam apps at industrial scale, consumer trust becomes even more valuable.

The Uncertainty Premium

Regardless of its flaws, Citrini's memo serves as a useful thought experiment. The fact that it rattled global markets reveals something important about our current moment: even sophisticated investors are struggling to model AI's economic implications.

This uncertainty creates both risks and opportunities. Companies that assume AI will simply automate away costs without changing competitive dynamics may find themselves blindsided. Those that prepare for more fundamental shifts in how value gets created and captured may thrive.

The memo also highlights a crucial policy challenge. If AI does generate massive productivity gains, how should those benefits be distributed? Market mechanisms alone may not ensure broad-based prosperity.