The ‘Prosperity’ Playbook

• Are we mistaking ‘short-term disruption’ for ‘long-term decline’?
• Will fear of job loss blind us to waves of better work history keeps delivering?
• Do we want policy that ‘performs’ — or policy that merely ‘feels good’?
• Will we throttle a general-purpose technology at birth, or let it compound human potential?
• And if the only window into the future is the past, why aren’t we looking through it properly?

AI is not a prophecy of doom; it is a productivity engine. The record is unambiguous: when technology collides with human ingenuity and free enterprise, living standards rise, opportunity widens, and the worst jobs disappear while better ones appear. The Industrial Revolution transformed a stagnant medieval economy into a society of durable abundance. The Digital Revolution repeated the act — at higher speed and scale. A third act is here. AI can, and will, bring prosperity — if we stop confusing transition pains with destination outcomes and allow markets to do what they do best: allocate talent to higher-value work.

History’s Scorecard: Disruption ‘Hurts’; Progress ‘Heals’

The Industrial Revolution unleashed, in barely a century, “more massive and colossal productive forces than have all preceding generations together,” as even Karl Marx conceded. In the U.K., between 1840 and 1900, real wages doubled. The average lifespan climbed by 22% — from roughly 41 to 50 years. Population doubled; employment rose by 80%. That isn’t a story of social collapse. It is a portrait of broad-based uplift.

Across the Atlantic, America experienced growth of biblical scale. From 1870 to 1900, real GDP tripled; population and labor force roughly doubled; manufacturing output grew sixfold. From 1865 to 1910, real per capita income rose 110%, while manufacturing workers’ real wages surged an estimated 173%. Life expectancy rose by a quarter as inflation-adjusted costs of food, clothing, and shelter fell by ~50%. Industrialization did not merely create jobs; it democratized comfort.

The Digital Revolution continued the pattern. Over the last quarter-century, U.S. real GDP rose by 66%. The Bureau of Labor Statistics reveals the American economy’s shock-absorber: since 2000, roughly 5.0 million Americans have separated from jobs each month — but the economy created ~5.1 million better-paying jobs a month. That churn is creative destruction at work. In 1810, 81% of Americans worked in agriculture; today, ~1.2% do — while America grows far more food, far more efficiently. And even as real industrial production quadrupled, the manufacturing share of the labor force fell to ~7.8% in 2025. Fewer people doing the same or greater output is not decline; it is productivity — and productivity is prosperity.

Yes, the transitions were painful. Luddites smashed looms. Intellectuals decried the age. Arnold Toynbee called the Industrial Revolution “disastrous and terrible.” Henry George claimed the rich grew richer while “the middle class is swept away.” But dire narratives consistently underestimated the adaptive capacity of free people in free markets. The outcome — measured by wages, life expectancy, and costs of necessities — refutes the fatalism.

Today’s Debate: Fear Sells, But It Doesn’t Build

Our era is replaying an old script. Some leaders propose taxes on robots or stipulate that AI “must” be bent to unrelated policy ends — unionization, environmental objectives, or identity-based mandates. Large-scale antitrust salvos aim broadly at “big tech,” risking collateral damage to the innovation flywheel. Labor leaders warn that technology is deployed to “eliminate workers or undermine and exploit us.” And a chorus calls for expansive “transition support,” guaranteed incomes, and lengthy benefit regimes that — despite good intentions — can slow re-employment and blunt incentives to reskill, delaying the very transitions that raise incomes.

Here’s the paradox: the transition costs are real, but policies designed to freeze the present make them bigger and longer. Europe’s stringent layoff restrictions often constrain job creation. China’s industrial subsidies prop up non-competitive sectors rather than future winners. American exceptionalism — higher productivity, higher living standards — has always rested on the willingness to adjust fast.

AI thrives in adaptive systems. It excels where talent can move, capital can flow, and experimentation can scale without permission-choked bottlenecks. Drag coefficients — overbroad licensing regimes, precautionary bans, political litmus tests for algorithms — don’t merely slow AI; they compound inequality by sheltering incumbents and starving challengers.

What AI Actually Changes: The ‘Job,’ Not the ‘Human’

AI automates tasks, not purpose. It compresses time-to-insight, reduces error, and elevates the floor of what entry-level contributors can do. Historically, when a general-purpose technology boosts productivity in one domain, it expands demand elsewhere: lower costs create new markets; new markets create new roles. The “old job” contracts; the “better job” — more judgment, more creativity, more human interface — expands.

• In agriculture, mechanization ended back-breaking toil and created value chains in logistics, processing, bio-sciences, and ag-tech.
   
• In manufacturing, automation shifted labor from repetitive assembly to quality, design, supply orchestration, maintenance, software, and customer success.
   
• In services, the internet reallocated attention from clerical drudgery to analysis, storytelling, and decision support.

AI is the next rung. It will erase the worst hours of work (rework, reconciliation, rote drafting) and amplify the best (strategy, problem framing, relationship building). The companies — and countries — that win will be those that reassign people up the value curve faster than rivals.

The Prosperity Operating System: Seven Principles

If we want the gains of history without needlessly magnifying the pains, we need a coherent, pro-growth framework. Here’s the playbook.

1. Permissionless Innovation by Default
   Regulate uses and harms, not the existence of the technology. Don’t pre-tax productivity (e.g., “robot taxes”). Treat AI like electricity: indispensable infrastructure governed by safety, liability, and sectoral rules, not blanket ideology.
    
2. Race-to-Reskill, Not Race-to-Replace
   Redirect funding from passive transfers to active upskilling: employer tax credits for skills-verified, AI-adjacent credentials; portable “skill wallets” that workers carry across employers; fast-track, outcome-tied bootcamps aligned to real vacancies. Make it easier to hire a newly trained worker than to write another automation script.
   
   
3. Productivity-First Antitrust
   Scrutinize conduct that forecloses markets; don’t punish scale that lowers prices and accelerates innovation. Encourage interoperability and data mobility so challengers can plug in and compete.
    
4. Talent Fluidity
   Reduce barriers to switching fields: sunset occupational licenses that don’t implicate safety; broaden apprenticeships; expand “earn-while-you-learn” pathways. Every week saved in reallocation is compounded lifetime income.
   
   
5. Outcome-Based Social Insurance
   Reform benefits to encourage rapid re-entry: wage insurance that tops up income for a time when workers take a lower-paid role to pivot; benefits that step down as skills step up; portability tied to the worker, not the job.
    
6. Public Data as Growth Capital
   Open high-value public datasets (with privacy safeguards) for model training and evaluation in health, transportation, energy, and education. When the commons improves, private solutions multiply.
    
7. Government as Lead Customer, Not Lead Scold
   Procure AI for service delivery — claims processing, fraud detection, permitting, tutoring — against clear KPIs (cycle time, accuracy, citizen satisfaction). Nothing de-risks technology for the private sector like a credible customer with hard metrics.

Answering the Objections — Directly

• “AI will kill more jobs than it creates.”
  History disagrees. The agricultural share fell from 81% (1810) to ~1.2% today, yet employment and output soared. In manufacturing, the labor share slid to ~7.8% by 2025, while real industrial production quadrupled. Since 2000, the U.S. labor market created ~5.1 million better-paying jobs each month while 5 million separated — net upgrading, not net erasure.
   
• “But this time is different.”
  Every time is different in detail; every time is similar in mechanism. General-purpose technologies reprice tasks. People and firms reallocate. Policy can speed or slow the reallocation. The prosperity delta is policy-dependent, not fate-bound.
   
• “We need guaranteed income to soften the blow.”
  Support must be bridges, not anchors. Extended, unconditional transfers have historically impeded transitions. Tie aid to skills acquisition, placement, and performance; design glide paths that pull people back into creation, not park them on the sidelines.
   
• “Big tech will monopolize AI.”
  Guardrails matter. But we should distinguish between scale that lowers cost and expands access and conduct that blocks competition. Interoperability, data portability, and targeted enforcement against exclusionary practices foster rivalry without kneecapping the productivity engine.

Where AI Lifts First — And Fastest

• Healthcare: AI accelerates diagnostics, triage, and drug discovery. The payoff isn’t merely cost cuts — it’s years of life.
   
• Education: Adaptive tutors can double teacher leverage, delivering personalized practice while freeing teachers for higher-order mentoring.
   
• SMB Productivity: Copilots collapse the gap between small teams and big back offices — bookkeeping, marketing, procurement, customer support — leveling the playing field.
   
• Public Services: Case handling, benefits adjudication, and fraud analytics become faster, fairer, and cheaper, restoring trust through responsiveness.
  
  
• R&D & Advanced Manufacturing: Generative design and simulation reduce time-to-prototype; quality systems catch defects upstream; supply chains become smarter and shock-resilient.

Each domain creates complementary roles — implementation, oversight, ethics, safety, change management, and, crucially, human relationship work. The more digital our workflows, the more human the differentiator.

Leadership Imperatives for CEOs & Ministers Alike

• Set a “70/20/10” AI Portfolio: 70% on core process uplift, 20% on adjacent growth bets, 10% on moonshots.
  
  
• Measure What Matters: Cycle time, error rate, customer NPS, cost-to-serve — tie bonuses to verified productivity gains.
  
  
• Codify Redeployment: Pre-commit to retrain-to-retain pathways; publish internal skills marketplaces; make upskilling a promotion path, not a side quest.
  
  
• Design for Dignity: Eliminate drudgery first. Use AI to elevate human judgment, not surveil it.
  
  
• Policy Clarity: Legislate use-based safety, data rights, and procurement reform — avoid vague “principles” that chill investment without improving outcomes.

The Stakes: A Second Economic Miracle

The market system, when allowed to absorb new technology, has achieved what no benevolent king’s redistribution, no church’s charity, and no mercantilist’s protectionism ever did: a massive expansion in productive capacity that compounds across generations. If we base AI transition policy on ‘proven results’ rather than ‘good intentions’ — flexible labor markets, vigorous competition, outcome-based safety, and relentless reskilling — we can author a ‘second economic miracle’ that enriches not just one nation, but the world.

History is shouting, not whispering. Progress punishes panic but rewards adaptation. The choice is ours.

Final Thoughts: Prosperity by Permissionless Progress

Fear is cheap; prosperity is earned. The last two centuries show a repeated pattern: technology displaces tasks, not human worth; markets reassign talent upward; living standards climb. Anchor policy to what works — flexibility, competition, reskilling, targeted safety — and AI will do what great technologies always do: create more, better, and fairer opportunity than it destroys. The future doesn’t need protection from AI. People need permission to build with it.

Image (c) istock.com