The Brontë Sisters Lost Money on Railways BUT the Railways Still Changed Everything

In 1845, everybody wanted a piece of the railways. Charles Darwin, John Stuart Mill and the Brontë sisters were all investing.

Within a year, most of them had lost money. Within a generation, railways had transformed Britain more completely than anything since the printing press.

Those two facts are not contradictory and there may just be a clue here for investing in AI mania.

A very quick history …

The Liverpool to Manchester railway opened in 1830 and was an immediate success. Dividends hit 7% that was the 'proof of concept' everyone needed.

By the mid-1840s, interest rates were low and railway shares could be bought on a 10% deposit — you paid the rest later. That one detail opened the door to almost everyone. In 1846 alone, Parliament passed 272 Acts authorising new railway companies. One issue of The Railway Times ran over 80 pages of railway investment ads.

Then costs came in higher than forecast, revenues lower, and the remaining 90% on those shares was called in. Stocks fell 50%. A third of the lines were never built. Families who'd staked their savings were wiped out.

And yet…

Within a generation railways had created seaside resorts, national newspapers, suburban housing, the telegraph, a hotel industry. Whole categories of business that couldn't exist before the railways became mainstream.

The people who got rich weren't mostly the ones who'd bought railway shares. They were the ones who built quietly on top of the infrastructure the mania left behind.

Three things I take from this:

The technology was real, the valuations massively over-optimistic. The bubble wasn't about a technology that failed — it was about 272 companies chasing markets that could support perhaps 30.

The mania built the infrastructure that made everything else possible. The lines that were built (even the money-losing ones) laid track that lasted 150 years. The AI equivalent is the LLMs, the chip innovation, the compute buildout. Most of it won't generate the returns investors are expecting. The AI market is already looking like a commodity market (in May for example DeepSeek's share of tokens jumped from 1 to 17% among developers on Vercel's platform). But if the infrastructure remains, the businesses built on top of it could be the ones that win.

The Brontë sisters were wrong about which company to back but that doesn't mean the technology didn't have potential.

The best question to ask is not who wins, but what can I do with the cheap technology that wasn't possible before?

Post 9of the "Every Era Has Its Reboot" series. | redebuter.com

Sources: Odlyzko (2010), University of Minnesota; Federal Reserve Bank of New York, Crisis Chronicles (2015). Vercel DeepSeek enters the fight for token volume, Anthropic continues to dominate spend, June 8th 20206