Thoughts on Artificial Intelligence Ahead of The World Government Summit

I am delighted to be joining the World Government Summit (WGS) 2025 this month, where, for three days, global leaders, policymakers and innovators, will meet to discuss, amongst other things, the transformative impact of artificial intelligence (AI) on governance, economics, and society. The conversations comes at a time when AI is advancing at an unprecedented pace, raising both great opportunities and profound challenges.

I've been reflecting on this topic and below are my thoughts in brief!

One of the most pressing concerns we must address is the cost-benefit equation of AI deployment. The recent emergence of China’s DeepSeek, a GPT-4-class model built at a fraction of the cost and computational demand of its Western counterparts, has exposed a critical flaw in the AI development race—our collective tendency to over-invest in AI without carefully considering its true return on investment.

As AI adoption accelerates, there is an urgent need for rational decision-making and financial prudence. Without a careful balance, we risk chasing an AI mirage - sinking vast sums of capital into inefficient systems while neglecting other critical areas of technological and economic growth.

AI Hype vs. Economic Reality: Are We Spending Wisely?

Over the past few decades, I have witnessed several cycles of technological overhype - from the dot-com boom to the blockchain and the NFT craze. Each time, we saw capital flood into overinflated expectations, followed by a sharp market correction that left many investors burned.

AI is on a similar trajectory, with massive investments pouring into generative models despite growing concerns about their long-term sustainability. Companies often justify AI investments by labeling them as ‘strategic,’ even when they fail to present a viable business case or a clear competitive advantage.

The development of DeepSeek, in China, demonstrates that AI models do not need to be built at an astronomical cost to be effective. This raises an uncomfortable question: Are companies unnecessarily overspending on AI projects simply to keep pace with the hype? I’ve seen it during the Y2K transition where CEO where ordering their CIO’s to spend more as the markets were judging the readiness of companies by the amount of money they spent on Y2K.

To ensure AI serves as a productive tool rather than a financial black hole, we must shift from indiscriminate spending to cost-efficient, purpose-driven AI deployment. The future of AI should be about efficiency, not extravagance.

The Contradiction Between AI and Climate Goals

Beyond financial concerns, AI’s energy consumption is on a collision course with global sustainability efforts. Governments worldwide have committed to reducing energy consumption under COP climate agreements, yet the current trajectory of AI development suggests the opposite; an exponential increase in energy demand.

Current AI models require immense computational power to function, with each new generation consuming twice the energy of its predecessor. AI data centers are already contributing to global energy shortages and increasing carbon emissions, a challenge that will only grow unless we rethink AI infrastructure. What I find hard to grasp is how can we, on one hand, push for aggressive climate action while, on the other, invest in AI systems that double our energy consumption? If we are truly committed to sustainable innovation, then AI development must be optimised for efficiency. The technology we build should be leaner, smarter, and more energy-conscious, rather than blindly following a trajectory that contradicts global sustainability efforts.

Thinking Revolutionarily, Acting Evolutionarily!

One of the fundamental mistakes businesses and governments make when adopting emerging technologies is rushing to deploy them. History shows that when a disruptive technology emerges, expectations skyrocket, capital floods in, and over-investment leads to inevitable correction. AI is following this pattern. The solution? Think revolutionarily, but act evolutionarily. I believe that;

? AI should be implemented incrementally, ensuring that investments align with measurable, tangible benefits.

? Instead of pursuing AI for the sake of AI, organisations must focus on use cases that provide clear returns and improve efficiency.

? AI should serve business goals, not dictate them. In other words the tail should not wag the organisation’s dog! Strategic alignment between AI and corporate objectives is more important than chasing the latest trend.

By taking an evolutionary approach, companies can minimise risk, maximise value, and avoid the pitfalls of technological overhype.

Next Leap: Cognitive AI and the Future Beyond Generative AI

While today’s AI breakthroughs are centered around generative AI, the next frontier is Cognitive AI, a model that moves beyond pattern recognition and integrates reasoning, adaptability, and contextual understanding. Unlike generative AI, which relies on vast datasets and statistical approximations, Cognitive AI introduces structured reasoning capabilities, allowing for:

? Context-aware decision-making

? Adaptability to new environments with minimal training

? The ability to explain and justify outputs

Generative AI lacks true intelligence - it predicts words, but it does not think. Cognitive AI represents a potential paradigm shift, moving AI closer to human-like decision-making and eliminating the inefficiencies of data-centric AI models.

As we invest in the next generation of AI, prioritising Cognitive AI over brute-force generative models could be a defining factor in AI’s long-term sustainability and impact.

Efficiency as the New Benchmark for AI Development

The development of DeepSeek in China has set a new benchmark for AI efficiency, proving that powerful AI models can be built with far fewer resources. This should prompt a fundamental shift in how we evaluate AI projects.

Rather than competing on size, compute power, or data consumption, the AI industry should compete on efficiency. Key questions AI developers should ask before building their next model:

? Can we achieve the same outcome with less computational power?

? Can we reduce energy consumption without compromising AI performance?

? Can we build AI systems that prioritize reasoning over brute-force data processing?

I started my computing career in an era where such thinking is essential. You got fired if you were wasteful with computer resources! AI should no longer be measured by how much capital is invested but rather by how efficiently it delivers value.

The Path Forward: AI Must Serve Humanity, Not Drain It!

As we gather at the World Governments Summit 2025, the conversation around AI must shift from raw technological ambition to efficient purpose-driven deployment.

The key questions for global leaders:

1. Are we investing in AI responsibly, or are we chasing an illusion?

2. How can we align AI development with climate goals rather than contradict them?

3. What role does Cognitive AI play in ensuring a more efficient, reasoning-based future?

4. How can we avoid the historical mistakes of overhype, overspending, and eventual market correction?

AI holds immense potential, but only if we develop, deploy, and govern it wisely. The future of AI must be one of efficiency, sustainability, and foresight. - not just an arms race of unchecked spending and energy consumption.

Regulation: Too Early, Too Costly, and Potentially Harmful

Another critical aspect of this discussion is the growing call for AI regulation. While ethical concerns and governance frameworks are important, we must recognise that regulating a technology that is still evolving could stifle progress, increase costs, and create unnecessary bureaucratic barriers.

Throughout history, premature regulation has slowed innovation rather than safeguarded it. Regulating AI today, while we are still discovering its full potential, risks setting arbitrary constraints that could make AI development prohibitively expensive and limit future breakthroughs. Worse still, heavy-handed regulations will likely benefit only the largest corporations that can afford compliance, while smaller players and innovators struggle under the burden of bureaucratic red tape.

Rather than rushing into rigid AI regulations, we need adaptive, flexible governance frameworks that evolve alongside the technology. The best approach is to focus on guiding principles rather than restrictive laws, allowing AI to develop organically, efficiently, and responsibly without curbing its potential before it has truly matured.

At WGS 2025, I look forward to discussing how we can develop AI systems that are not just powerful, but purposeful. AI should be a tool for progress, not a burden on our economies and planet. Now is the time to get AI right.

Jamil El-Imad

CC Christophe Reech Elizabeth Fox Manuel Blanco Kristalina Georgieva Jasem Albudaiwi Omar Sultan AlOlama Abdulla Mohammed Al Basti