Why the U.S. Military Keeps Investing in Yesterday’s Weapons

Last week, I found myself reading about mounted cavalry units in World War II. Yes, you read that right—horses. The British and American militaries literally kept cavalry regiments on the books well into the 1940s, even as tanks were rolling across Europe. It sounds absurd from where we sit today, but it made me think about something that still happens in military spending: we often pour billions into weapons systems that are becoming obsolete before they ever see real combat.

The parallels between those cavalry units and today’s military procurement challenges are striking. And honestly? They matter way more than you might think.

The Cavalry Problem Never Really Went Away

Here’s what happened with mounted cavalry: military leaders had invested decades into cavalry traditions, training, and doctrine. Generals who’d built their entire careers around horses weren’t exactly eager to admit the game had changed. Even with tank technology emerging, they argued that cavalry served specialized purposes—reconnaissance, rapid deployment, flexibility. Sound familiar?

The thing is, this wasn’t stupidity. It was institutional momentum mixed with genuine uncertainty. When new technology emerges, it’s not always obvious how to use it or whether it’ll actually work in real conditions. So organizations cling to what they know.

Fast forward to now, and we’re seeing the same pattern play out with military technology spending. The U.S. defense budget is massive—we’re talking hundreds of billions annually—but a chunk of that still goes toward legacy systems that might not be the right tools for modern conflicts. Drones, cyber warfare, and rapid-response AI systems are increasingly shaping real military scenarios, yet funding still flows heavily toward traditional platforms that made sense ten years ago.

When Tradition Becomes Expensive

What fascinates me is how hard it is to stop funding something once it’s started. You have:

• Contractors with factories and thousands of employees
• Military leaders who trained on these systems
• Congressional districts that depend on the jobs
• Genuine sunk costs that make cancellation feel wasteful

Take a hypothetical weapons program that costs $50 billion to develop. By year ten, you’ve spent $40 billion and you’re starting to realize there’s a better approach. Do you kill it and lose that $40 billion, or finish it and spend $10 more? Politicians and generals often choose the latter, even if the former makes more sense.

I’m not saying this is always wrong. Sometimes you need to finish what you started. But sometimes you’re just throwing good money after bad, which is exactly what happened with cavalry units. They kept funding those regiments right up until they met actual tanks in actual combat. Then they discovered, rather conclusively, that horses don’t work against armored warfare.

The Real Problem: Guessing About the Future

Here’s the uncomfortable truth nobody likes to admit: we’re not very good at predicting what weapons systems will actually matter. Thirty years ago, nobody knew drones would become central to modern warfare. Fifteen years ago, cyber attacks seemed like a niche concern. Today, we’re probably making similar mistakes about what will matter in 2035.

The difference is that military spending moves slowly. A weapons system approved today won’t reach full operation for a decade or more. By then, the threat you were preparing for might have changed completely.

What we need is more flexibility. More willingness to kill failing projects. More investment in emerging technologies rather than protecting established ones. And more honest conversations about what we actually need versus what we’re comfortable with.

The cavalry units were eventually disbanded, but not until after they’d cost time, money, and military credibility. We have a chance to learn from that mistake. Whether we actually will? That’s the real question.

Source: The U.S. is Blowing Billions on the Wrong Weapons