For most of human history, the concept of sustained economic growth was a fantasy. Before 1700, the global economy essentially stagnated, with output creeping up by a mere 0.1% per year on average. The Industrial Revolution changed everything, quintupling growth to 0.5% annually by 1820. Today, Silicon Valley evangelists are predicting an even more dramatic shift. They argue that Artificial General Intelligence (AGI)—AI capable of outperforming humans at most cognitive tasks—could soon propel annual GDP growth to a staggering 20-30% or more.
The Engine of Explosive Growth: Ideas, Not Just People
Historically, economic expansion was tied to population growth. More people meant more labour, but also, crucially, more ideas. These ideas eventually led to higher productivity and rising living standards. AGI presents a paradigm shift: it promises runaway innovation without a corresponding increase in human population, potentially supercharging GDP per person.
While economists agree AI can boost productivity, the scale is hotly debated. Some, like Daron Acemoglu of MIT, project a modest total GDP boost of 1-2% over a decade. This view assumes only about 5% of tasks can be automated more cheaply by AI. However, this research is based on 2023-era AI capabilities. More radical projections assume that as AGI is achieved, a far larger share of economic output will be automated. The constraint then shifts from human labour to energy, infrastructure, and capital investment.
Beyond Automation: The Self-Improving AI Loop
True explosive growth requires AI to master the ultimate task: improving technology itself. If AGI agents can conduct scientific research, make novel discoveries, and enhance their own capabilities, we enter the realm of "endogenous" growth theory. Here, ideas beget more ideas at an accelerating pace. Research group Epoch AI models a scenario where, once AI automates about a third of tasks, annual GDP growth surpasses 20% and keeps rising.
This could lead towards an economic "singularity," a theoretical point of infinite growth. While likely a logical extreme, even a significant acceleration would be transformative. The critical question is what this means for the workforce.
Labour, Wages, and the Baumol Effect in an AI World
The first Industrial Revolution did not immediately benefit workers; real wages stagnated for centuries. In an AGI-driven boom, the fear is not stagnation but redundancy. The cost of running an AGI would cap wages, pushing them down over time. Eventually, all income could accrue to capital owners.
However, economist Tyler Cowen notes that real-world constraints—energy, regulation, data, or institutional sluggishness—may slow this transition. If humans retain an edge in some areas, they will work alongside machines. Some, particularly "superstar" workers, could see explosive wages, even if labour's share of the total economy shrinks.
For others in automatable desk jobs, the outlook is tougher. They might shift to sectors resistant to automation, like skilled trades or personal services. These sectors could experience "Baumol's cost disease," where wages rise due to demand from a wealthy AI-owning class, even without productivity gains. This creates a dual economy: abundance in AI-produced goods (cheap food, digital entertainment) but high costs for labour-intensive services like childcare or dining out.
Investing for an AI-Powered Future: A High-Stakes Puzzle
If you believe explosive growth is coming, the models offer clear, if challenging, advice: own capital. But knowing which assets to hold is complex. Breakneck growth would likely trigger sky-high real interest rates, as identified by economist Frank Ramsey. Consumers, expecting vast future wealth, would save less, forcing rates up to incentivise investment.
This creates a tug-of-war for asset prices like stocks. High rates depress valuations, but high growth boosts future earnings. The net effect is ambiguous. Land, with fixed supply, could become valuable for data centres and energy infrastructure, but is highly sensitive to interest rates. Cash deposits could benefit from high rates, but risk inflation if central banks misjudge the situation.
Governments with high debt would face a squeeze: growth helps their finances, but soaring borrowing costs hurt. This could intensify demands for redistribution, like universal basic income. Despite the buzz, financial markets are not yet pricing in this explosive growth scenario. Bond yields have even fallen after major AI announcements, suggesting skepticism.
Yet, AI has consistently defied capability forecasts. If its economic impact is similarly underestimated, investors and policymakers alike are in for a seismic surprise. As economist Robert Lucas noted, the consequences of growth for human welfare are so profound they eclipse other concerns. AGI places that observation on steroids, forcing us to rethink the very foundations of our economic future.
