The Algorithm of War: How America and Israel Built the World’s Fastest Targeting Machine

 

Satellites, AI, and software now hunt targets faster than humans can think. America and Israel have built a digital war machine that can unleash hundreds of strikes before the world even understands what happened.

War used to move like a freight train. Loud, slow, grinding forward mile by mile. Today it moves like software—silent, fast, and terrifyingly precise. I have watched the transformation unfold, and the truth is blunt: America and Israel have built vast military targeting machines powered by software and artificial intelligence. These systems do not merely help armies fight wars. They supercharge the process of finding things to bomb.

Let me call a spade a spade. War has gone industrial again.

When the United States and Israel unleashed their campaign against Iran, the tempo shocked even hardened military observers. On February 28 the allies flew more offensive strike sorties in a single day than America managed on the opening day of major combat during the 1991 Gulf War or the 2003 invasion of Iraq, when about 1,300 sorties were launched. Five days later U.S. war secretary Pete Hegseth bragged that Operation Epic Fury delivered twice the air power of the famous “shock and awe” campaign that stunned Baghdad in 2003.

I can almost hear the old generals from the twentieth century shaking their heads. War used to depend on pilots scanning maps and analysts poring over grainy photographs. Now data pours in from satellites, drones, intercepted signals, and even social media. The battlefield has become a digital river, and the armies that control the data control the kill chain. Inside U.S. Central Command headquarters in Tampa, Florida, the process runs like a war factory. Intelligence officers from the J2 directorate assemble a massive database containing thousands of potential targets. Satellite imagery shows buildings. Electronic signals reveal military radios. Drone feeds track vehicles. Analysts combine it all into a living map of destruction.

Schools, hospitals, and protected sites are placed on “no strike” lists. Everything else becomes a potential target.

Once the database exists, another specialist enters the picture—the weaponeer. This officer matches weapons to targets the way a mechanic chooses tools. Deep bunker? Use a bunker-buster bomb. Hardened building? A GPS-guided Joint Direct Attack Munition, known as a JDAM, will do the job. Missile launcher spotted on a road? A drone strike might be quicker.

A lawyer reviews the decision, but the role is limited. A former American commander once explained the reality with brutal honesty. The lawyer does not say you cannot do it. The lawyer says you can do it, but here are the consequences. Then the commander decides. All of this used to take hours or days. Now software compresses the process into minutes.

The secret weapon is the Maven Smart System, a powerful decision-support platform developed largely by Palantir Technologies. Maven gathers information from every possible source. Open-source data flows in from the internet. Classified intelligence arrives from satellites and electronic surveillance. The system fuses it together into a single digital picture.

Imagine an Iranian posting on Telegram that he saw a missile launcher passing through his neighborhood. Maven captures the message. At the same moment a radio-frequency satellite detects the electronic chatter of Iranian military radios in the same area. The software links the two clues, marks the location, and suggests a target.

That is how the machine works.

NATO officer Arnel David, who oversees aspects of the program, describes the ambition in almost scientific terms. The goal is to turn military command into a “machine-aided predictive science.” War planners can simulate the blast of a weapon, estimate the radius of heat and fragmentation, and project civilian risk before a missile ever leaves the launcher.

On a computer screen the blast zone appears as a jagged shape soldiers call a “splat.” Inside that splat, everything dies.

The speed of the system borders on the surreal. Joe O’Callaghan, a retired U.S. Army colonel who helped develop Maven, once revealed that planning a war-scale operation that previously required dozens of staff working for tens of hours could now be done with roughly one-tenth the manpower. A former NATO general put the transformation even more bluntly. Tasks that once required hours can now be executed in about two minutes.

The number of targets generated each day has exploded. One European general described the change with a single phrase: “alchemy.” The old pace was about 10 targets per day. Today it can reach 300. The aspiration is 3,000. That is not strategy. That is industrial production.

Israel has built its own version of the system. Its origins stretch back to the trauma of the 1973 Yom Kippur War, when Israeli aircraft were savaged by Soviet-built air defenses. Israeli intelligence responded by building a sophisticated system for mapping enemy surface-to-air missile sites. The result was a devastating victory over Syrian defenses in 1982. The concept expanded over time. By the 2006 Lebanon war against Hizbullah, Israeli commanders discovered a new problem. Their aircraft were flying so many missions that they ran out of targets. One frustrated general reportedly complained that the target bank had been exhausted.

Military intelligence chief Amos Yadlin solved the problem by expanding the database. Soon Israel possessed a massive catalog of potential targets across hostile territories—bases, command centers, weapons factories, and militant hideouts.

The modern Israeli system now operates with astonishing speed. When rockets launch from Gaza, commanders can open a binder—or more accurately a digital dashboard—select a target, and retaliate within minutes. That is the power of algorithmic warfare.

But machines make mistakes, and war punishes mistakes brutally. On February 28 a Tomahawk cruise missile struck a girls’ school in Minab, southern Iran. 175 people died, most of them children. U.S. officials later said the missile had been aimed at a nearby naval base and that the tragedy resulted from faulty targeting data. By early March the human-rights organization HRANA reported nearly 1,800 deaths in Iran, many of them civilians.

Critics argue that turning war into a high-speed data pipeline risks catastrophic errors. Some NATO officers worry about the gradual erosion of human control. Artificial intelligence models such as Claude, developed by Anthropic, are already used in limited ways within defense systems. For now they do not identify geographic targets directly. But the debate has begun. How much authority should a machine have in war? An Israeli intelligence officer once explained the dilemma with brutal simplicity. Artificial intelligence can make a good officer better. But if an officer is careless, the machine will simply help him find more targets faster.

That truth haunts every digital battlefield. The deeper danger lies in scale. When computers generate hundreds of targets per day, human analysts struggle to keep up. Buildings that once housed militants may now contain families. Intelligence data ages quickly. A failure to revalidate targets can turn yesterday’s military site into today’s civilian tragedy.

Meanwhile the political climate shifts. Secretary Pete Hegseth has emphasized “lethality” over what he calls “tepid legality.” Civilian-harm assessment teams inside the Pentagon have reportedly been cut by 90%. Within CENTCOM some planning cells operate with only one-third of their previous staff.

The result is a strange and unsettling paradox. The targeting machines are smarter than ever. Yet the human oversight around them may be shrinking.

I sometimes think of an old proverb while watching this new era unfold. When the sword grows sharper, the hand holding it must grow wiser. Right now the sword is becoming frighteningly sharp.

America and Israel have built targeting systems that can turn oceans of data into bombs within minutes. Satellites whisper, algorithms calculate, and missiles fly. The battlefield is no longer measured only in miles. It is measured in milliseconds.

War has entered the age of software.

And software never sleeps.

 

For readers interested in a separate line of thought, the titles in my “Brief Book Series” are available on Google Play. Read them here on Google Play: Brief Book Series.