A VC Argues Data Centers Are the Railroads of Our Era

A venture capitalist is betting that today's data center boom rhymes with the steel ships and railroads of the Industrial Revolution — and Fortune has published the full argument.

The piece, which ran under the headline "Steel ships, railroads, and data centers: A VC takes the Industrial Revolution analogy seriously," makes a claim that sounds like conference-stage shorthand and then actually pursues it. The investor's framing is simple: the physical infrastructure of the 1800s — steel hulls, rail lines, the machines of mass production — may be the closest historical parallel to the compute infrastructure now being built at scale.

That comparison has moved from cocktail talk to capital allocation. Data centers have become one of the largest infrastructure build-outs in modern corporate spending, and the Fortune story gives the analogy a serious editorial platform rather than treating it as a throwaway line.

Why does the Industrial Revolution comparison matter?

The historical pattern the VC is invoking runs through three phases. First came the hardware — steel ships that shrank distance, railroads that moved goods and people at unprecedented speed. Then came the infrastructure layered on top: depots, ports, distribution networks. Finally came the businesses that couldn't have existed without the first two layers.

The modern mapping is direct. Chips and servers play the role of steel and steam. Data centers are the railheads and ports. And whatever gets built on top of that compute layer — the applications and companies still being formed — is the part historians will write about later.

The takeaway in plain words: the money is going into the rails, because nobody knows yet which trains will run on them.

What does the analogy get right — and where does it strain?

The strength of the comparison is capital intensity. Railroads and steel shipping required enormous upfront spending before a single unit of revenue came back, which is exactly the cost profile of large-scale compute facilities today. Investors who funded the railways often didn't capture the value; the operators and customers who used the network frequently did.

That's the uncomfortable implication for anyone pouring money into physical AI infrastructure right now. If the pattern holds, the biggest winners of the compute era may not be the ones laying the concrete.

The analogy also carries a warning about excess. The railway manias of the 19th century produced spectacular booms and equally spectacular busts, with overbuilt lines and ruined speculators, even as the underlying infrastructure permanently changed the economy. A VC endorsing the parallel is implicitly endorsing both halves of that history.

What should investors watch next?

The Fortune piece lands at a moment when the analogy is doing real work in boardrooms. If data centers truly are the railroads of this cycle, the questions become concrete fast: How much capacity gets built? Who finances it? And who ends up owning the profitable layer once the build-out matures?

History's answer to the last question was rarely the ones who built the tracks — and that's the number every infrastructure investor should be checking against their own thesis.

By Grace Kim