By Mohamed Ali - Technology & Digital Economy Commentary
May 19, 2026 - The S&P 500 returned 31% between May 2025 and May 2026. That’s a strong year by almost any measure. But that’s not the real story. The real story is what happened beneath the index itself. Chip companies. Data center infrastructure. Energy systems feeding AI compute. That layer didn’t return 31%. It exploded.
Chip companies averaged returns of more than 320%. Data center and energy infrastructure names averaged over 419%. Some individual companies posted even sharper gains. Bloom Energy surged 1,647%. Micron climbed 770%. Intel rose 483%.
If you were positioned in what I call the “Singularity Loop” , chips, infrastructure, and energy became one of the market’s clear winners. Investors positioned there didn’t just beat the broader market. They outpaced it repeatedly. The concept itself is relatively simple.
As AI systems become more capable, they require exponentially more computing power. That creates rising demand for advanced chips. Those chips require massive data centers. Those facilities require enormous amounts of electricity. Every layer feeds the next. And as demand for AI scales globally, every link in that chain becomes more valuable.
The Infrastructure Layer Powering AI#
The structure resembles earlier technology transitions. When mobile networks expanded globally in the early 2000s, several layers of infrastructure benefited simultaneously: telecom towers, hardware manufacturers, and the software ecosystem built on top. The difference with AI is speed.
The cycle is compressing into a much shorter timeframe, while attracting levels of capital rarely seen in previous technology waves. According to the Stanford HAI 2026 AI Index Report, global corporate AI investment reached $581.7 billion in 2025, representing a 130% increase year over year.
The momentum accelerated even further in early 2026. The first quarter of the year reportedly broke venture capital records, with approximately $300 billion deployed globally in a single quarter. Around 80% of that capital, roughly $242 billion , flowed into AI-related companies.
That matters because this is no longer being treated as speculative experimentation. Capital markets are increasingly treating AI infrastructure as a foundational economic layer. Chips have effectively become the physical substrate of intelligence. Without them, large-scale AI systems cannot function. But chips alone are not enough.
Training and operating advanced AI models requires hyperscale data centers built specifically for compute-intensive workloads. Those facilities must be cooled, connected, and powered continuously.
And power itself is rapidly becoming one of the defining bottlenecks of the AI era.
A single advanced AI training cluster can consume electricity on the scale of a small city. That reality is pushing renewed interest toward nuclear energy, grid-scale storage, clean gas infrastructure, and alternative energy systems capable of supporting AI demand at scale.
The result is a feedback loop where AI growth increases the value of the infrastructure supporting it. That loop is already reshaping capital allocation globally.
The Gap Between the Old Economy and the New One#
The contrast between AI infrastructure assets and traditional sectors is becoming increasingly difficult to ignore. Real estate reportedly returned 5% during the same period. Financials returned 11%. Healthcare gained around 9%.
Those are not weak numbers.
Related articles
But they exist in a completely different category from the returns generated by AI-linked infrastructure companies. Even the broader technology sector, which returned approximately 76%, significantly lagged behind the companies sitting directly inside the infrastructure layer powering AI.
The gap reflects something larger than sector performance. Global capital appears to be reorganizing itself around the systems enabling artificial intelligence. Every economy has a finite amount of investable attention. Right now, a growing share of that attention is moving toward one thing: the infrastructure running AI systems.
The long-term projections reinforce the direction of travel. Research firm Gartner projects global AI spending could reach $2.52 trillion in 2026 alone. PwC estimates workers with AI skills already earn significantly more than peers without them. Broader forecasts suggest AI could add trillions of dollars to global GDP over the coming decade.
At this stage, the numbers are no longer just speculative forecasts. They are signals from capital already in motion.
The AI Labs Building the Engine#
At the center of the ecosystem are frontier AI labs building increasingly advanced models. Most remain privately held, which means ordinary investors cannot directly buy them. But their valuations reveal how aggressively markets are pricing the future of AI. Over the past year:
- OpenAI reportedly rose from a $300 billion valuation to $852 billion.
- Anthropic moved from approximately $61.5 billion to nearly $1 trillion.
- xAI climbed from $80 billion to $250 billion.
- Mistral AI increased from $6.2 billion to $14 billion.
These figures are based on private funding rounds rather than audited public valuations. Still, they reflect real capital being deployed by investors making concentrated bets on the future of AI systems.
Anthropic’s reported Q1 2026 funding round alone reached $30 billion. OpenAI’s reportedly hit $122 billion, making it one of the largest private funding rounds in history. But the larger story may not be the labs themselves.
Every trillion-dollar AI company requires a multi-trillion-dollar supply chain beneath it. Semiconductors. Compute infrastructure. Cooling systems. Energy generation. Grid capacity.
That supply chain is publicly traded. And increasingly, that appears to be where much of the long-term value accumulation is happening.
Africa Is Already Inside This Story#
One of the biggest misconceptions around AI is that it remains concentrated inside Silicon Valley. It doesn’t. Research from KPMG suggests that many of the countries with the highest rates of day-to-day AI adoption are emerging economies, including Nigeria, India, Egypt, and Brazil.
That matters because the tools being built by trillion-dollar AI labs are already being used daily across cities like Nairobi, Lagos, Cairo, Mogadishu, and Accra. The infrastructure value may still be concentrated elsewhere, but the adoption layer is becoming deeply global. The pattern itself is not new.
The people who built telecom infrastructure in the 1990s looked early before the mobile revolution arrived. The people who understood data centers before cloud computing scaled looked early too.
The same cycle now appears to be repeating around AI infrastructure. This is not necessarily about buying a specific stock tomorrow morning. It is about understanding where generational wealth and long-term economic leverage may be accumulating during the current technology transition. The Singularity Economy is already running. The loop is already paying. The remaining question is who understands the infrastructure layer early enough to position around it.