I have found myself taking way too long to write posts, so I am editing less and going faster. Hopefully, quality and coherence don’t suffer too much.
I have been fortunate enough to have worked in tech earlier in my career. Most of the projects and companies I worked in weren’t successful. We often built cool software that we were convinced would really help the businesses and customers we were focused on serving, but the problem was that the customers often didn’t sign up and buy what we were selling. Sometimes it was marketing and sometimes it was because we just hadn’t done enough work understanding our target customers.
This goes back to a tried-and-tested rule of business failure that Silicon Valley thinks they coined, but didn’t: “if you build it, they will come.” If you don’t have a clear set of customers who are demanding your product or service, you are spending time and capital on a mere hope. You’re assuming that if you build it, the customers will just flock to your product.
This hope is what is powering the financing of A.I. businesses. Hope not data. Nearly $100 billion was invested in AI infrastructure in the last quarter of 2025, and estimates are that over $2 trillion will be invested by the end of 2026. This is eerily reminiscent of one of the great failures in American business history: the implosion of the railroad bubbles of the postbellum United States, which began with a banking crisis in 1873 and handicapped the American economy for two decades.
And “built it, and they will come” is—outside of corn-nested Iowa baseball fields where it totally works—not a sensible business strategy. The tech companies, whose stocks go up, and the financiers and banks that fund whose fees go up, are all clearly aligned in this effort. Their aligned interests make sense. They both get money. Unfortunately, though, they aren’t happy with just feeding off of each other. They have decided to hedge their bets by taking the economy hostage and enlisting politicians on both sides to do it.
Banks and hedge funds have taken the economy hostage through the sheer size and scope of the investments being made. Banks and hedge funds are throwing hundreds of billions of dollars into these companies—many of which, it must be said, will not be successful—and they’re doing so because they’re convinced that they will not be left to collapse if they fail. The 2008 financial crisis and the politics that followed—Obama, Trump, and Biden—did not hold bankers or companies accountable for their bad bets. The moral hazard does not seem to be a concern for anyone. And bankers and large investors know it. Sprinkle in the ability for companies to pay politicians to support them via SuperPACs and paid speeches and cushy post-Congress jobs, and you have a compromised state that enables the hostage-takers.
The hostage is the whole economy, and politicians keep arming the hostage taker. Both Democrats (not quite all of them) and Republicans (most of them, I’d say) are bullish on A.I. Partly, this is driven by a genuine belief in the power of this technology (which, by the way, I share.) But that’s not the only reason politicians are boosting A.I.
Inflation is up, unemployment is strangely emergent in places (now amongst white collar workers) we haven’t seen perhaps ever, and the war in Iran is damaging just about every market in every place on Earth. American politicians, particularly those in power—that is, Republicans right now—are boosting A.I. because they know it is the only growth that can keep the numbers that they use to claim success looking good. A.I. is a sugar rush to make up for systemic problems.
We have to stop re-arming the hostage-taker. Now. As great as A.I. technology is—and I am most definitely not anti-A.I. I believe this technology can bring real wealth to our economies and real improvements to people’s lives. The railroad did both of those things, too, but that doesn’t mean the bubbles, corruption, and insider trading it enabled didn’t bring the American economy to its knees. It did.
Politicians will eventually form a new coalition to attack this corruption and develop this technology sensibly and carefully, without relying on hope over careful logic. People did it in the 1870s and ‘80s, and they will come together to do so again. The question is how severe the social pain will have to be for people to stop enabling this over-investment and start protecting our economic system. It got really bad in the 1870s, and eventually led to the first civil service reforms and anti-corruption crackdowns.
Hopefully, we don’t have to have another twenty-year stagnation to get back to that.
Leave a Reply