Answer extracted from the Proven Podcast — listen to the full episode below.
Why U.S. capitalism and post-AI economies are fundamentally misaligned
Current U.S. capitalism is built on requiring 2% GDP growth, 4% unemployment, and a standard work model from age 22 to 65 with eight-hour days and two weeks vacation—a framework that becomes structurally impossible in a post-AI economy. The core tension is that society must accelerate AI development for military and competitive advantage while simultaneously redesigning the economy to function without traditional employment as the primary income model.
The economic model underlying modern capitalism assumes perpetual growth and continuous labor demand. This model depends on workers remaining in the market for 43 years, with unemployment holding steady around 4% to maintain equilibrium. But artificial intelligence disrupts every assumption behind this structure.
As Mark Roberge explains in the episode, the incompatibility isn't theoretical—it's baked into the immediate policy paradox we face. Governments and enterprises recognize that AI adoption must accelerate rapidly for competitive and military reasons. Yet accelerating AI adoption directly undermines the labor-dependent economic model that governments are trying to sustain.
The collision between acceleration and adaptation
The United States ranks 22nd globally on happiness indices and is declining—a signal that the current structure is already straining. The tension between needing to advance AI and needing to maintain employment is not something that can be resolved through minor policy tweaks or retraining programs alone.
A post-AI world will not support work schedules where the same proportion of the population is continuously employed for four decades. If AI delivers on its productivity promises—with metrics like rep-to-manager ratios expanding from 7-to-1 to 15-to-1—then the total number of jobs available structurally contracts. The economic growth model that has relied on employment as the primary wealth distribution mechanism cannot function unchanged.
"Product-market fit is not about selling people, getting revenue, closing customers—it's about the customers realizing the value you promised, and that is best quantified by retention."
Mark Roberge — Managing Director at Stage Two Capital and Professor at Harvard Business School. Roberge was the fourth employee at HubSpot, joining as the first salesperson and serving as founding CRO through the IPO over a nine-year period ending in 2013. He was recruited to join Harvard Business School's faculty 13 years ago to teach sales to the MBA program. Eight years ago, Bessemer Venture Partners approached him to start Stage Two Capital, the first VC firm run by the best sales and marketing leaders in tech, which has raised four funds deployed across 150 startups.
Beyond economics, there's a deeper question about how society structures meaning and purpose when continuous employment ceases to be the default path. This reframing is discussed in detail in the full conversation, where Roberge explores not just the numbers but the philosophical shift required.
See also
How does AI adoption compare to previous major technological and societal transitions?
The AI movement is not comparable to the transition to the internet but rather to fundamental shifts like when humanity moved from nomadic to agricultural societies and from agricultural to industrial ones.
What is your definition of product-market fit and why is the traditional definition flawed?
Product-market fit is not about selling people, getting revenue, or closing customers—it is about customers realizing the value you promised, best quantified by retention.
How should organizations measure AI enablement in their sales teams in 2026?
The two input measures are selling time—the percentage of a seller's week spent with a buyer or prospect, which should increase from 25-30% to 75%—and the rep-to-manager ratio, historically at 7-to-1 in tech but targetable to 15-to-1 with AI enablement.
Key takeaways
- U.S. capitalism requires steady 2% GDP growth and 4% unemployment, built on a work model from age 22 to 65—assumptions that break down with widespread AI adoption.
- The core paradox is that governments and enterprises must accelerate AI for competitive and military advantage while simultaneously redesigning the economy to function without traditional employment models.
- A post-AI economy cannot sustain the same proportion of the population in continuous, full-time employment for 43 years, requiring fundamental rethinking of how wealth is distributed and meaning is structured.
- The urgency of this incompatibility is already visible: the United States ranks 22nd on global happiness indices and is declining, signaling strain in the existing system.