If AI continues to grow, NVIDIA won’t be the only one to win. #aistockswhatnext
When we think about the AI boom, we normally think about NVIDIA, AMD, or the big companies developing models.
But there’s something easy to forget: AI doesn’t work in thin air.
It needs chips.
It needs memory.
It needs servers.
It needs data centers.
It needs electricity.
It needs cooling.
It needs networking capable of moving enormous amounts of information.
And while everyone is watching the stock performance of the companies that make the chips, an entire economy is being built around them.
In fact, NVIDIA is paying increasing attention specifically to that infrastructure. This week, it unveiled a program for power and cooling products aimed at its “AI factories,” and it’s also working with Google and Emerald AI on systems that allow data centers to adapt their electricity usage to grid conditions.
I find this interesting because it changes the question.
It’s no longer just:
“How far can AI stocks go?”
We can also ask:
“What does AI need in order to keep growing?”
And then completely different sectors come into view: energy, cooling, networking, storage, data centers, and computing.
But there’s another layer I’m even more interested in: Web3.
If AI needs more and more computing capacity, storage, infrastructure, and payment systems, some decentralized networks might try to become part of that infrastructure.
That’s where concepts like DePIN, decentralized computing, storage, AI agents, and on-chain payments come in.
The interesting question is a different one:
Which projects are truly solving part of this infrastructure—and capturing value from it?
Maybe the next big AI-related opportunity isn’t about manufacturing the next chip; maybe it’s about everything that makes it possible for those chips to work.
Which part of this chain do you think we’re still underestimating?
When we think about the AI boom, we normally think about NVIDIA, AMD, or the big companies developing models.
But there’s something easy to forget: AI doesn’t work in thin air.
It needs chips.
It needs memory.
It needs servers.
It needs data centers.
It needs electricity.
It needs cooling.
It needs networking capable of moving enormous amounts of information.
And while everyone is watching the stock performance of the companies that make the chips, an entire economy is being built around them.
In fact, NVIDIA is paying increasing attention specifically to that infrastructure. This week, it unveiled a program for power and cooling products aimed at its “AI factories,” and it’s also working with Google and Emerald AI on systems that allow data centers to adapt their electricity usage to grid conditions.
I find this interesting because it changes the question.
It’s no longer just:
“How far can AI stocks go?”
We can also ask:
“What does AI need in order to keep growing?”
And then completely different sectors come into view: energy, cooling, networking, storage, data centers, and computing.
But there’s another layer I’m even more interested in: Web3.
If AI needs more and more computing capacity, storage, infrastructure, and payment systems, some decentralized networks might try to become part of that infrastructure.
That’s where concepts like DePIN, decentralized computing, storage, AI agents, and on-chain payments come in.
The interesting question is a different one:
Which projects are truly solving part of this infrastructure—and capturing value from it?
Maybe the next big AI-related opportunity isn’t about manufacturing the next chip; maybe it’s about everything that makes it possible for those chips to work.
Which part of this chain do you think we’re still underestimating?