Why AI’s Growth Curve Could Outpace the COVID Shock to the Global Economy

A graph that reframes the AI conversation

A single graph circulating among technology researchers is reshaping how experts think about artificial intelligence growth. According to analysis by research firm Metr, AI capability is currently doubling roughly every seven months, a pace that suggests its economic and social impact could rival, or even surpass, the disruption caused by the COVID pandemic.

Unlike pandemic driven shocks, which were sudden and externally imposed, AI’s expansion is unfolding through sustained and compounding technological progress. This difference matters, because it points to a transformation that builds over time rather than peaking and fading.

Understanding what is actually doubling

The growth highlighted by Metr is not about hype or user adoption alone. It focuses on effective compute, a measure that combines raw processing power, algorithmic efficiency, and training optimisation. When these factors are combined, the result is a steep curve that shows AI systems becoming dramatically more capable in a short period.

Metr researchers estimate that there may be at least another five years of this rapid doubling cycle. If that projection holds, the capabilities of AI systems in the early 2030s would be orders of magnitude beyond what is available today.

This kind of exponential growth is rare outside of foundational technological shifts such as electrification or the rise of the internet.

Why comparisons to COVID are emerging

COVID reshaped economies almost overnight. Supply chains broke, labour markets shifted, and governments were forced into emergency interventions. AI’s trajectory is different, but potentially broader.

Rather than disrupting one sector at a time, AI touches almost every knowledge based activity simultaneously. From software development and finance to healthcare and creative industries, AI tools are increasingly embedded into daily workflows.

The key similarity lies in scale. Just as COVID forced institutions to adapt faster than expected, AI may compel businesses, regulators, and workers to rethink systems that were designed for a slower pace of change.

Productivity gains and economic tension

Supporters of AI growth argue that rapid capability gains will unlock major productivity improvements. Tasks that once required teams of specialists can increasingly be handled by smaller groups using AI tools, lowering costs and accelerating output.

However, these gains also create tension. Productivity does not automatically translate into shared prosperity. Without policy adaptation, benefits may concentrate among firms that control data, infrastructure, and capital, while labour displacement accelerates elsewhere.

This dynamic mirrors early phases of previous industrial revolutions, where efficiency gains outpaced social adjustment.

Why the next five years matter most

The next half decade is critical because institutions tend to lag technology. Legal frameworks, education systems, and labour protections are often calibrated for linear change, not exponential curves.

If AI capability continues doubling every seven months, systems built around multi year planning cycles may struggle to keep up. Businesses that delay adoption risk falling behind competitors, while governments that delay regulation risk reacting too late.

The speed itself becomes the challenge, not just the technology.

Limits and uncertainty remain

Despite the steep curve, uncertainty remains. Physical constraints, energy costs, and data availability could slow progress. Researchers also caution that capability growth does not automatically mean alignment with human values or safe deployment.

There is also the question of diminishing returns. While current gains are dramatic, future improvements may face harder problems that are not solved simply by more compute.

Still, even a partial continuation of the current trend would represent a historic shift.

A structural shift rather than a moment

The most important insight from the Metr graph is that AI is not a single breakthrough moment. It is a structural shift unfolding at compound speed.

COVID compressed years of change into months. AI may compress decades of technological evolution into a few years. The difference is that this change is being built deliberately, not imposed by crisis.

That makes preparation possible, but only if institutions accept the scale of what is coming.

Why the graph matters

Graphs matter because they anchor abstract ideas in measurable reality. In this case, the curve suggests that AI’s impact will not be incremental or easily absorbed.

If capacity continues to double on its current path, the question will not be whether AI reshapes economies and societies, but how well they adapt before the curve steepens further.