Welcome to 2026, where the Pentagon has finally achieved its ultimate dream: turning international diplomacy into a high-stakes game of Fruit Ninja. According to Katrina Manson’s latest book, *Project Maven: A Marine Colonel, His Team, and the Dawn of AI Warfare*, the US military has officially “learned to love AI.” And by “love,” they mean using the Maven Smart System to hit over 1,000 targets in Iran within 24 hours. Because nothing says “innovation” like doubling the efficiency of a “Shock and Awe” campaign while your coffee is still warm.

Let’s take a moment to appreciate the sheer, unbridled logic here: that hitting twice as many things, twice as fast, is inherently “better.” It’s the Amazon Prime-ification of kinetic warfare. Why wait three weeks for a strategic assessment when an algorithm can swipe right on a thousand coordinates before the afternoon briefing? The assumption that speed equals success is a bold one, particularly when the “targets” are identified by computer vision software that started its life trying to tell the difference between a golden retriever and a loaf of bread. We’ve gone from “Don’t Be Evil” to “Don’t Be Slow,” and the results are apparently explosive.

The narrative arc of Project Maven is truly a touching “underdog” story—if the underdog was a multi-billion-dollar defense apparatus. Manson’s investigation highlights Maven’s humble beginnings in 2017 as a scrappy experiment in drone footage analysis. Remember when Google employees protested the project, causing a massive corporate existential crisis? How quaint. In the high-speed landscape of 2026, those ethical concerns feel like a dial-up modem in a Starlink world. The military didn’t solve the “human-in-the-loop” problem; they just shortened the loop until the human is basically a glorified “Accept Terms and Conditions” button.

The claim that Maven has “taught” the military to love AI suggests a romantic comedy where the Pentagon is the grumpy bachelor and the algorithm is the manic pixie dream girl with a penchant for high-resolution thermal imaging. But let’s be real: the military loves AI because it offers the ultimate bureaucratic shield—the “Algos made me do it” defense. If you level 1,000 targets in a day, you’re not making tactical decisions; you’re trusting a black box that’s been trained on datasets we’re told are “robust,” but which can still be fooled by a strategically placed piece of cardboard or a particularly dusty lens.

Furthermore, the obsession with the “scale” of the Iran assault—nearly double that of the 2003 Iraq invasion—is a fascinating metric for progress. In any other industry, doubling your output without doubling your understanding of the consequences is called a “product recall.” In defense tech, it’s a “dawn of a new era.” We are operating on the assumption that more data leads to better targeting, ignoring the fact that AI is famously prone to “hallucinations.” It’s one thing when ChatGPT hallucinates a fake legal precedent; it’s quite another when a targeting system hallucinates a legitimate military objective in a crowded urban center.

As we celebrate the “success” of Maven, we should probably ask if we’ve actually improved warfare or if we’ve just automated the chaos. The 2026 reality is a world where the speed of war has officially outpaced the speed of human thought. We’ve traded strategic nuance for algorithmic efficiency, proving that while you can teach a general to love an AI, you can’t necessarily teach an AI to understand the geopolitical mess it’s cleaning up—or creating—at 1,000 targets per day. But hey, at least the SEO on the mission reports will be fantastic.