Pakistan and the price of waiting

“I feel behind”. These were the words of Andrej Karpathy, one of the architects of modern AI, speaking last week.

Andrej Karpathy is not an observer of technological change. He is one of its architects. A man who helped build Tesla’s Autopilot, co-founded OpenAI and trained some of the most capable engineers of our time admits he cannot keep up.

This is not uncertainty or fatigue. It is a leader acknowledging that the pace of technological change has outstripped even those who built it. Boris Cherny, principal engineer at Meta for seven years and now leading Claude Code at Anthropic, noted, "I start approaching a problem manually, then have to remind myself, Claude can probably do this”.

When those who built the foundations of modern technology speak in such terms, it is a warning to the world. If even the architects feel outpaced, what does that say about nations still debating whether innovation matters?

In many emerging economies, time is not just slipping but being taken. While others set the rules, build the systems and claim leverage, hesitation becomes surrender. Delay is no longer neutral. It is a decision to forfeit influence, control, and economic power to those who move faster. The global stage is being written in real time and those who arrive late will inherit systems they did not build.

In 2011, I discovered bitcoin. I read the code, traced every line and felt the tremor of a revolution. I understood its power – and yet I waited. I told myself it was early. I told myself to be cautious. By the time I acted, the future had already moved without me. Knowledge alone was never enough. Courage to act is what defines those who shape history.

Watching today’s wave of innovation, I feel the same sting of deja vu but on a far greater scale. Signals are blazing and impossible to ignore. Most have noticed. Few have truly understood. Across much of the developing world, there is growing recognition of the importance of technological progress.. Awareness alone is no longer sufficient. History does not forgive hesitation. It devours those who wait while others build, innovate and seize the future.

The era of the junior developer as we knew it is quietly ending. Software is no longer built line by line. Today, engineers increasingly describe outcomes rather than instructions. Systems generate code, test it, refine it and deploy it. The developer’s role is shifting from construction to orchestration, managing intelligent systems, agents, workflows and feedback loops.

Boris Cherny recently spent weeks working without opening a traditional development environment. The models did the work. He supervised. Yet in many emerging economies, the system still prepares students for roles that are rapidly disappearing. Talent is ready. Opportunity is lagging. The gap is growing. The junior developer is dead. The future belongs to those who can act decisively and shape the story before it is written for them.

The scale of investment and impact is staggering. Entire industries, economies and nations are being rewritten in real time: McKinsey estimates that innovation could add $4.4 trillion annually to the global economy. Goldman Sachs reports that nearly 300 million jobs may be affected worldwide. Meta is investing $35 billion in infrastructure in 2024. Microsoft and OpenAI project $100 billion in investment this decade. NVIDIA reported $47 billion in data centre revenue last year.

This is not corporate competition. This is national infrastructure on the scale of railways, electricity grids and nuclear capability. Countries failing to build will not merely fall behind. They will import decisions, ideas and productivity. They will surrender strategic autonomy. Most policy discussions miss this. Analysis, prediction and execution are accelerating faster than decision-making frameworks. Tools can already generate solutions faster than humans. Decision-making and responsibility remain uniquely human.

The problem is not the scarcity of options but the scarcity of judgment, courage and commitment. AI cannot choose. It cannot bear responsibility. As capabilities rise, the burden of human decision intensifies. Those who act decisively and responsibly will define the future. Those who hesitate will inherit systems shaped by others.

Many governments announce strategies, form committees and host conferences. Real preparedness rests on three pillars: compute capacity; technical talent; and institutional speed. Across much of the developing world, gaps remain across all three pillars. Advanced infrastructure is mostly imported. Domestic research is fragmented. Public-sector decision-making moves slower than progress itself. Technology accelerates while institutions remain immature. Capability rises faster than the structures meant to guide it.

In 2024, global AI investment highlights the widening divide: the US invested $109 billion, 24 times more than the UK, the third-largest investor. Pakistan is only beginning to step into AI, moving cautiously as the world accelerates. Among developing nations, only India and China attract meaningful investment. We are not merely falling behind. We are spectators to our own erosion. Every advancement occurs elsewhere while we debate and delay. Our children will inherit systems we did not build. Our economy will be shaped by hands we cannot control. This is a slow national surrender.

Pakistan produces thousands of computer science graduates each year, yet few experience the scale and depth of modern innovation: high-performance resources remain limited; training in collaboration and strategic thinking is rare As a result, talent departs abroad or remains idle. Ambition without opportunity breeds frustration, and talent without infrastructure leads to a slow bleed of national capability.

These are fractures weakening the nation, leaving the next generation to inherit a world they did not shape. The solution is clear. Build world-class AI hubs locally, combining cutting-edge infrastructure with mentorship in strategic innovation. Empower talent to stay, create and lead, turning ambition into national capability and giving Pakistan a seat at the table where the future is being written.

Emerging economies need execution. Priority must be given to national facilities capable of powering world-class research; public-private hubs to harness collective innovation; regulatory frameworks that encourage experimentation; programmes to retain and repatriate talent; and governance that keeps pace with change. Authority today is earned through prioritising amid abundance, excluding plausible alternatives, acting without guarantees and accepting consequences openly. The future will not be shaped by those who predict best, but by those willing to take responsibility for decisions that cannot be proven correct in advance. Dario Amodei, CEO of Anthropic, warns that the risk is not power. It is the immaturity of institutions to wield it responsibly.

When Karpathy says he feels behind, and I watch my country struggle in the same way, we are facing the same crisis from two perspectives: one personal, one national. In 2011, I ignored bitcoin. I understood its signals but hesitated. By the time I acted, the opportunity had passed. I did not fail from ignorance. I failed from delay. Today, the stakes are far higher.

Karpathy’s words are a warning. When one of the architects of change feels outpaced, it shows how far ahead the frontier has moved and how far behind we already are. This is not alarmism. It is a call for clarity, courage and decisive action. Pakistan still has a window, but it is narrowing. Every day of hesitation deepens dependence and erodes influence. The question is no longer whether we understand the stakes. It is whether we are willing to act decisively, responsibly and now.

Beyond national policy, the private sector plays an equally decisive role. Every major corporation and growing business must now ask a simple question: how is artificial intelligence improving our efficiency, decision-making and competitiveness? AI adoption can no longer be treated as an IT upgrade. It is a strategic function.

One of the most practical steps forward is for serious companies to appoint a chief AI officer or equivalent role – ideally a young, technically fluent builder who understands modern tools, experimentation and rapid iteration. Someone embedded in leadership, not isolated in a back office. In many cases, this individual will be under 30, fluent in emerging development workflows, comfortable with automation and focused on execution rather than theory. Their mandate should be clear: redesign processes, remove friction and turn technology into a measurable advantage.

For forward-looking firms, this may prove to be one of the highest-return investments they ever make. Not in software, but in institutional adaptability. In an era where intelligence is becoming a core input of production, companies that fail to integrate it at the leadership level will steadily lose relevance, regardless of their past success.

Arnold Toynbee warned, "Civilisations die not by murder, but by suicide”. Hesitation, inaction and failure to adapt is the suicide of nations. History does not honour those who planned to adapt. It remembers those who built the rails while others debated the route.

The writer is the chairman of the Pakistan Virtual Assets Regulatory Authority (PVARA).

Disclaimer: The viewpoints expressed in this piece are the writer's own and don't necessarily reflect Geo.tv's editorial policy.

Originally published in The News