Every generation likes to believe it is living at the peak of human advancement — but the numbers tell a quieter, more worrying story. Over the past four decades, the world has been getting better, yes, but it has also been getting slower at getting better. Economic growth has softened, scientific output has plateaued in many fields, productivity has flatlined in advanced economies, and regional inequalities have widened again after decades of narrowing. It is a strange moment: we possess more technology than at any point in history, yet the engine of progress feels as though it’s coughing.
Consider what the long-term data shows. Global Total Factor Productivity, which economists treat as a proxy for innovation, grew rapidly from the 1950s to the 1990s. Then something changed. From 2005 to 2020, productivity growth in OECD countries hovered around 0.5% — levels not seen since the Great Depression era. Likewise, global GDP per capita still rises, but at a slower rate than during the explosive catch-up years of the 1990s and early 2000s. Even human development indicators, such as gains in health and education, have begun to plateau, especially after 2015. Many charts that once curved upward now bend sideways.
This stagnation narrative is not coming from pessimists. It appears in publications such as The Economist, in academic work from Harvard, MIT, and Stanford, and in long-run data series from the World Bank, UNDP, IHME, and OECD. The message is consistent: progress has not stopped — but its pulse has weakened.
And it is precisely here that the conversation about AI becomes complicated. For the first time in decades, we have a technology that seems big enough — system-changing enough — to potentially jolt the world into a new era of growth. But AI’s arrival has collided with an uncomfortable truth: global economies are entering this technological shift from a weakened position, not a strengthened one.
This is why the current debate feels tense. On one hand, AI shows extraordinary promise. It compresses time, amplifies human capability, and lowers the cost of intelligence-driven work. But on the other hand, AI enters a world grappling with falling birth rates, declining scientific productivity, rising geopolitical fragmentation, and a labour market that has barely recovered from the waves of outsourcing and automation already behind us.
And so, a key question emerges: will AI revive progress, or merely displace the little progress left?
Much depends on whether AI becomes a multiplier or a substitute. If AI multiplies human work — making the average person far more productive — we could see the first serious uptick in global productivity in more than 20 years. But if AI mostly substitutes for human labour — amplifying the “jobless growth” pattern already present in many economies — then we may face a future in which the technology races ahead. At the same time, wages, living standards, and opportunities stagnate.
There are reasons to be cautious. Studies from the IMF, McKinsey, and the ILO estimate that between 30% and 45% of global work tasks could be automated or significantly altered by AI. Historically, societies absorb such shocks when the underlying economy is strong. But today’s baseline is fragile. Large segments of the global workforce are already mismatched, underemployed, or ageing out. Governments are fiscally stretched. Education systems are slow to adapt—and capital increasingly pools at the top.
Yet there is also reason not to despair. Throughout history, periods of stagnation have been followed by bursts of ingenuity. The electrification era came after a prolonged productivity slowdown. The internet boom arrived after the sluggish 1970s. Often, progress appears to stall right before it transforms. The question is whether we are now on the cusp of another transformation — or another deepening slowdown.
AI could be the catalyst that breaks the cycle. But only if we use it as a lever to expand human capability, not compress human opportunity. Only if it broadens access to knowledge rather than concentrating power. Only if it accelerates scientific discovery, productivity, and inclusive growth rather than amplifying the world’s existing inequalities.
The world is not doomed to slow progress. But the data is clear: the trajectory of improvement we once took for granted has weakened. AI arrives not as a silver bullet, but as a provocation — a reminder that technology alone does not guarantee progress. How societies choose to integrate, govern, and distribute the gains of AI will determine whether we experience a new uplift in human advancement or a deeper phase of stagnation.
In the end, the real story is not that progress has stalled or that AI will save or ruin us. The real story is simpler and more human: we are entering a moment when the future is, once again, undetermined. And that makes this one of the most important — and genuinely exciting — chapters in modern history.