The Productivity Question Nobody Is Asking: How Much AI Growth Do We Actually Want?

Since ChatGPT stormed into public consciousness in late 2022, economists, technologists, and business leaders have been locked in a familiar debate: how dramatically will generative AI transform productivity? But buried beneath the forecasts and the hype lies a more fundamental question — one that is as much philosophical as it is economic. How much AI-driven productivity growth should society actually want?

It is a question that Michael R. Strain, Director of Economic Policy Studies at the American Enterprise Institute, raised in a recent analysis, and it is one that resonates deeply within the automation and AI industry.

History Warns Us: Growth Has a Human Cost

The historical record on rapid technological change is sobering. During the British Industrial Revolution, average wages stagnated or declined for at least the last two decades of the 18th century. Workers were displaced. Cities swelled with people living in overcrowded, unsanitary conditions. Disease spread. Entire occupational categories were wiped out almost overnight.

It was this very upheaval that inspired Charles Dickens to chronicle Victorian poverty and Karl Marx to theorise the consequences of what he called "constantly revolutionizing the instruments of production." As Marx and Engels wrote in 1848, such change causes disruptions to "the whole relations of society" — a sentiment that was not abstract theorising but a direct observation of the world collapsing and rebuilding around them.

The warning is relevant today. As AI accelerates the automation of knowledge work, creative industries, customer service, and beyond, the risk of large-scale displacement is real. The question is not whether disruption will occur, but whether our institutions, safety nets, and education systems are equipped to manage it.

The Stakes Are Enormous — In Both Directions

The economic case for aggressive productivity growth is compelling. Productivity is the primary engine of long-term living standards. During the peak of the 1990s digital revolution, productivity growth reached approximately 3 percent annually — a rate that would double Americans' standard of living in roughly 24 years. By contrast, the sluggish 1.5 percent growth that followed the 2008 global financial crisis would take 47 years to achieve the same outcome.

Some analysts now suggest that AI could push productivity growth to 5 or 6 percent — nearly double the pace of the internet boom. At that rate, the material benefits in terms of new medicines, safer workplaces, longer lifespans, and greater leisure time would be profound. Future generations would look back at today's living standards the way we look back at the pre-electrification era: with disbelief at how difficult daily life once was.

But speed matters enormously. The faster the pace of technological change, the greater the social instability, and the harder it becomes for displaced workers to adapt and find new employment. This tension between the promise of transformation and the pain of transition sits at the heart of every serious conversation about AI adoption.

A Different Perspective: Automation as Empowerment, Not Just Efficiency

For Hamza Baig, founder of the Automation Institute and CEO of Hexona Systems, the debate is not merely academic. It is the foundation of his life's work.

"The real risk is not that AI grows too fast," said Baig. "The real risk is that we let it grow without bringing people along with it. Productivity without literacy creates a two-tier economy — those who operate the machines, and those who are operated on by them. Our mission is to make sure as many people as possible are on the right side of that line."

Baig's perspective reflects a growing school of thought that sees workforce education and automation literacy as the critical variable in determining whether the AI revolution mirrors the empowering trajectory of the late 20th-century information age or the damaging dislocation of 18th-century Britain.

Policy Must Keep Pace With Technology

Strain's analysis suggests one key reason the United States navigated the technological disruption of the 1990s better than Britain did in the 1700s was the presence of a mature social safety net and broader access to education. These structural supports allowed workers and communities to absorb change more effectively than their historical counterparts could.

The implication is clear: the outcome of the AI revolution will not be determined solely by the technology itself, but by the policy environment surrounding it. Retraining programmes, income support, access to affordable education, and thoughtful regulation of labour markets will all play defining roles in whether the gains of AI-driven productivity are broadly shared or narrowly concentrated.

Three key questions frame the policy challenge: What is the optimality criterion — maximum output, reduced poverty, or protection of the most vulnerable? How much weight should current workers' interests receive relative to the living standards of future generations who will inherit the benefits? And critically, can the pace of adoption be managed without choking off innovation entirely?

The Creative Destruction Bargain

Economic disruption has always come paired with creation. The communities devastated by manufacturing job losses in the American Rust Belt over the past several decades experienced very real and lasting harm. Yet few people today would genuinely choose to reverse the technological progress of the 20th century if given the option.

As Strain observes, the politics of nostalgia often overwhelm the economics of progress. But the historical record consistently shows that societies that embraced technological transformation — while investing in people — ultimately produced better outcomes for more citizens than those that resisted change or ignored its human costs.

The AI era will be no different. The question is not whether the transformation will happen. It will. The question is whether businesses, policymakers, and educators move quickly enough to ensure that the benefits of that transformation are accessible to the many, not just the technically privileged few.

For advocates like Hamza Baig, that work is already well underway.