Who Aims the Gun?

Military Advantage Shifts from Mass to Mind

A drone rises from a field, guided in real time by a human operator watching a live video feed. The pilot sees what the drone sees: a tank comes into view. The drone accelerates. The image fills the screen. Then: static.

A machine worth millions of dollars is taken out of battle by one costing a few thousand. For several years, this has seemed like the defining asymmetry of modern conflict. Cheap, disposable drones overwhelm expensive industrial platforms. Brains beat brawn. True enough—but incomplete.

A first-person-view (FPV) drone is not intelligent in itself. The intelligence sits in the operator. Every maneuver, every correction, every last-second adjustment depends on a human brain interpreting a video stream and issuing commands.

The weapon is cheap. The brains remain scarce. Scale is limited not only by manufacturing capacity, but by the availability of trained operators. One drone, one pilot. Saturation depends as much on manpower as on hardware. Not for long.

Increasingly, drones carry onboard processing. Machine-vision systems classify vehicles. Algorithms distinguish armored targets from decoys. Navigation routines compensate for jamming. Swarms coordinate without constant human direction. The intelligence migrates from the operator to the machine.

Now, one human can supervise many systems. The contest becomes less about how many drones are produced, and more about how quickly systems can detect, classify, decide, and act. In military theory, that cycle has a name: the decision loop. Whoever closes it fastest gains the advantage.

The same evolution holds for defense.

Directed-energy systems—high-powered lasers integrated into layered air defenses—promise to alter the economics of countering drones and rockets. Once deployed, a laser shot costs little more than electricity. There is no interceptor to replenish, no magazine to exhaust.

That economic advantage depends on information. A laser system must detect small objects against cluttered backgrounds. It must distinguish friend from foe. It must track erratic motion. It must decide which of multiple simultaneous threats to engage first.

The hardware matters. But the decisive layer is sensing, classification, and targeting—the informational architecture that allows the system to respond fast enough to matter.

Israel offers a vivid illustration. Its layered air-defense architecture—integrating Iron Dome, David’s Sling, and laser systems such as Iron Beam—continuously analyzes incoming trajectories, classifies threats, predicts impact points, and decides in seconds which projectiles must be intercepted, and which can be allowed to fall harmlessly.

Call it algorithmic triage. The system conserves expensive interceptors, reduces unnecessary engagements, and increasingly relies on real-time data fusion and AI-assisted targeting to compress the decision loop under combat conditions.

Even with AI assistance, drones remain material weapons operating in the physical realm. Laser defenses increasingly operate in the informational realm. And information systems improve faster.

In recent days, we have seen dramatic evidence of how far precision has advanced. Full casualty data from the joint U.S.-Israeli strikes on Iran is not yet available and much remains contested. But even early reporting suggests a continued migration toward highly targeted operations enabled by dense sensor networks, real-time data fusion, and advanced discrimination.

Remember when cruise missiles and early drones were hailed as “smart” alternatives to area bombardment? And they were. Yet even as late as the post-2014 U.S. campaign against ISIS, the U.S. military officially acknowledged more than 1,400 civilian deaths from coalition air operations in Iraq and Syria. Precision reduced harm relative to earlier eras—but it did not eliminate it.

The next stage in this evolution is not simply better explosives, but better discrimination.

Compared to even a decade ago, today’s precision-guided systems incorporate far denser sensor networks, faster data fusion, and increasingly algorithm-assisted targeting. While the full effects of recent campaigns remain incomplete, the trajectory is unmistakable: accuracy depends less on explosive yield and more on information dominance.

The more warfare becomes a problem of detection and classification, the more it rewards computational speed over industrial mass. For centuries, military advantage flowed from steel, fuel, and manufacturing scale. The drone era seemed to tilt the balance toward cheap material swarms overwhelming expensive material platforms.

Now that balance is shifting again—toward cognition compressing the decision loop, moving warfare deeper into the informational realm.

For investors, firms operating at the intersection of sensing, AI, targeting, electronic warfare, and integrated system coordination will become increasingly interesting. In a growing crowd of wannabes, investors must seek out those who close the loop fastest in harmony with physics.

At Gilder’s Technology Report, we look for technologies that exploit science to alter cost curves. When power once provided by mass is achieved instead by intelligence, entire sectors can be revalued.