You think you know what it means – but you don’t.

You know there’s a new railroad being built. It’s on the news every day. It’s been driving the stock market higher and higher. Hundreds of billions of dollars a year are going into data centers – computers the size of cathedrals, drawing electricity measured in gigawatts, rising out of farmland in Ohio and Wyoming and Texas.

This is the largest infrastructure build-out in America since the railroad boom of 1865-1873.

You know this will change your life and the lives of all Americans, in many ways. You know fortunes will be made, fortunes measured in trillions of dollars. But what you don’t know is: most investors are about to lose everything.

There’s a huge paradox at the heart of every technological revolution. Tech innovations destroy as much as they create. And it is this creative destruction that lies at the heart of capitalism’s ability to dynamically create vastly more wealth.

The most important thing for investors today isn’t to know what artificial intelligence (“AI”) will build. The most important thing is to understand what it cannot destroy.

Let me start by explaining the key economic driver of these changes: an increase in computer efficiency that’s hard to even imagine, increases that are driving the price of compute down to virtually zero.

Nvidia (NVDA) built GPU chips that perform thousands of small calculations simultaneously rather than a few large ones in sequence. Broadcom’s (AVGO) switching silicon and Nvidia’s own chip-to-chip links made it possible to move data between processors fast enough that 10 thousand separate chips behave like a single machine. High-bandwidth memory from SK Hynix (SKHY) and Micron Technology (MU) places memory directly against the processor, so the chip stops waiting for the information it needs.

Parallel processing gave us the horsepower. Bandwidth lets us harness thousands of engines to one plow. Memory stopped the engines’ idling. Together, they collapsed the cost of computation.