Today’s chart is deceptively simple, but it might be the most impactful chart you see this year.
Because a year ago, the biggest question around AI was whether it would live up to the hype.
But now the question is whether the world can keep up.
An Infrastructure Boom
Here’s this week’s chart.
Image: https://x.com/stocktalkweekly/status/2033622201988256162
As you can see, there’s not much to it.
One bar shows roughly $500 billion in demand. The next one doubles that number.
This photo was taken at Nvidia’s recent GTC conference in San Jose, CA, where CEO Jensen Huang made a stunning prediction that the world is headed toward at least $1 trillion in demand for AI computing infrastructure by 2027.
What makes this chart fascinating isn’t just the size of the number. It’s what has changed to make this jump possible.
Since the “ChatGPT moment” in late 2022, the main focus has been on who is building the best AI. In other words, who has the biggest model with the best benchmarks and the most advanced capabilities.
That phase was all about making AI practical.
What’s happening now is what makes it valuable.
Every time someone queries an AI system, every time a piece of software leans on a model to make a decision and every time an automated workflow runs in the background, it requires compute.
That’s where things have drastically changed over the last year.
Because AI systems are now being embedded into everyday software and workflows. The more useful it becomes, the more it’ll get used. And the more AI gets used, the more infrastructure it requires.
That’s why Microsoft’s AI business has already scaled into a multi-billion-dollar run rate. Meta is pouring tens of billions into infrastructure to support its own AI features. And Amazon is retooling AWS to handle a new class of workloads that didn’t exist just a few years ago.
But as I’ve written about before, these are operating expenses tied directly to products people are already using.
Which is why the “AI bubble” narrative misses something important.
There’s a reasonable argument that spending has gotten ahead of itself. We’ve seen similar cycles before, where infrastructure gets built out faster than demand materializes.
But the difference today is that the demand isn’t hypothetical. It’s already here.
And it’s only going in one direction.
Up.
Here’s My Take
It’s easy to look at a trillion-dollar forecast and assume it’s driven more by hype than reality.
But I don’t believe that this chart merely represents a surge in optimism from a CEO who stands to benefit from it.
I see it as a shift in how artificial intelligence fits into the economy. It shows me that demand is being pulled forward by usage, not pushed ahead by speculation.
Right now, AI is moving from something you experiment with to something that runs continuously in the background. In that way, it’s more like electricity than software.
And we’re still in the early stages of the global AI infrastructure buildout.
Which means, if Huang’s prediction is right, that trillion-dollar number isn’t a ceiling.
It’s merely a starting point.
Regards,
Ian King
Chief Strategist, Banyan Hill Publishing
Editor’s Note: We’d love to hear from you!
If you want to share your thoughts or suggestions about the Daily Disruptor, or if there are any specific topics you’d like us to cover, just send an email to dailydisruptor@banyanhill.com.
