Seventy miles south of Hanoi, the Thanh Hoa Bridge spanned the Song Ma River, seemingly invincible. The North Vietnamese called it the Dragon's Jaw, for good reason. Over the years, American fighter jets had flown 869 bombing raids on the bridge, losing 11 aircraft. After each mission, the smoke cleared, and the bridge still stood, a monument to the futility of aerial bombardment. Seven years of futility.

One morning in May 1972, the jets came again: F-4 Phantoms racing through the sky. Below their stubby wings: an experiment, a new bomb cooked up at Eglin Air Force Base in Florida's panhandle, rigged to follow a beam of laser light flashed from above.

One by one, 10 jets rolled in and let their bombs go. The bombs locked on to the laser beam illuminating the target and hit dead on, engulfing the bridge in flames. This time, when the smoke cleared, the pilots saw the bridge had been knocked clean off its 40-foot concrete abutment. The Thanh Hoa Bridge had fallen. Smart bombs had made their mark. War would never be the same.

If President Bush authorizes an invasion of Iraq in the weeks or months to come, within hours Iraq's air defenses will most likely be destroyed from close range by stealth aircraft the Iraqis can't see, and from long range by stealth missiles they can't stop. The opening salvo will be unprecedented, if not for its duration then for its lethality, as bombs, guided by sophisticated electronics, communicating directly with satellites, home in precisely and simultaneously on hundreds and hundreds of targets across the country, from presidential palaces in Tikrit to troops in the field near Basra.

Unmanned flying drones will circle high above the battlefield beaming continuous surveillance to commanders on the ground -- in real-time video, radar and infrared -- and to pilots in the air, enabling them to track and destroy moving targets minutes after they emerge from hiding. Bunker-busting bombs will pierce Iraq's most protected places, penetrating up to 100 feet into the earth, crashing through reinforced concrete, exploding at precisely the depth desired.

No other nation even approaches this new American military power. Not since 1945, when the United States had a brief monopoly on atomic weapons, has there been such a power gap between America and the rest of the world.

How did we get here? There was no equivalent of the Manhattan Project, when the biggest names in physics gathered in an all-out effort with the massive, unified support of an entire government. Instead, single, often unsung, individuals had eureka moments, connecting the dots of emerging technologies in unforeseen ways that changed the world. It took years, and increasingly impressive proof on the battlefield, before these inspirations were recognized for what they were -- a new way of fighting that would change the calculations of war and peace in unprecedented, and still uncertain, ways.

On a balmy autumn day in 1964, Air Force Col. Joe Davis Jr. watched from the low roof of an office building in Orlando as engineers at Martin Marietta, a defense contractor, demonstrated something called a laser. The laser looked like a couple of cigar boxes mounted on a tripod. The engineers shined the gizmo at a plywood target being towed along an elevated track about 2,000 feet away.

The laser was straight out of Buck Rogers, right up Davis's alley. He headed an Air Force unit scouting out new technology for the war in Vietnam. The laser was developed by an Army scientist who imagined that someday it might help guide anti-tank missiles. But Davis was a pilot, a decorated ace from World War II and the Korean War, and as he watched the laser dot stay focused on that target moving in the distance, something clicked.

"Boy," he said, "we ought to get this laser to steer our bombs."

Soon, he climbed into the back seat of a military training jet back at Eglin, a base two-thirds the size of Rhode Island, with enough open space for an army of Air Force officers and bomb scientists to test just about anything that could be dropped out of an airplane.

With a movie camera in hand, Davis directed the pilot to fly to a big dam and circle it. If his movie camera were a laser, could he hold it still enough from a moving aircraft to keep the laser beam fixed on a target?

"When we got back I had the photo lab develop the film," Davis recalls, "and hell, those cross hairs stayed right on that damn dam."

Davis knew he was on to something.

A few months later, an executive from Texas Instruments named Glenn E. Penisten showed up at Eglin hoping to sell the Air Force some technology. Penisten mentioned that he had a new ground-to-air missile that might possibly be rigged to follow a laser beam. Davis asked Penisten and his TI team out to dinner at a fancy seafood and steak restaurant in Fort Walton Beach. Over surf and turf, he leveled with them. "I've got a bombing problem," he said. American planes were dropping tons of them, but they weren't hitting the targets. Could a laser guide bombs dropped from aircraft? After a couple of drinks, Penisten recalls, they started sketching the first laser-guided bomb on the back of a napkin.

Not long after the dinner meeting, the TI team returned for further discussions. Engineer Weldon Word and several colleagues explained to Davis what they had in mind: a kit that could be attached to regular 500- and 2,000-pound bombs consisting of a laser sensor on a bomb's nose and movable fins on its tail. The sensor would be able to detect a laser light reflecting off a target.

The sensor's field of view was divided into quadrants -- whichever quadrant was stimulated by the laser sent an impulse to the bomb's tail fins, making them move up or down. The resulting change in the bomb's direction would push the laser into another quadrant of the sensor, triggering another movement of the tail fins. The net result was that whenever the bomb headed away from the target, the tail fins activated to keep the bomb on course.

Davis listened intently. It was a Friday afternoon. He told them that if by Monday morning they could come up with a proposal that would come in under $100,000, "I'll fund you."

So Word and the other TI engineers flew back to Dallas, stayed up all weekend and came back on Monday morning with a 12-page, handwritten proposal for building 12 laser-guided bombs in six months for $99,000. Davis looked over the proposal, then told Word and a colleague to come back after lunch.

Word and his friend didn't know what to expect. "There was 99 percent spit juice in that thing, and 2 percent fact," says Word, 71, now retired and living in Tyler, Tex. They were exhausted, and overwrought. They bought a six-pack of beer and slept on the beach outside the base.

They'd need their rest. Instead of giving them a contract outright, Davis's superiors in Air Force research and development at Wright Patterson Air Force Base in Ohio brought in another defense contractor, North American Aviation, and made it a competition. But that was only part of the problem. Word and company had no experience with lasers, and only a dim notion of how to build a laser sensor. Plus they had no idea whether the tail fins in their proposal, adopted from TI's Shrike missile, could control and stabilize something as heavy as a 2,000-pound bomb.

Lacking money for wind tunnel experiments, they built small 10-inch models of bombs with fins and dropped them into a backyard swimming pool, taking notes as they sank.

Nobody but Davis had much confidence in the idea. Penisten even remembers how his own boss bet him a case of Jack Daniel's that the laser-guided bomb would never work. But in relatively short order, it all came together in a kit that cost less than $2,000. All they needed was a real bomb to put it on. Penisten called Davis. Davis loaded a dummy bomb into the cargo bay of a C-47 and flew it to Dallas. Word still remembers the day it arrived at corporate headquarters. "All of a sudden word goes through the building -- there's a bomb on the back dock!"

Davis is 85 now, a short, thick man with a full head of silver hair and leathery, suntanned skin. He lives a mile and a half from Eglin's main gate with his wife, Ann. He spends his days golfing. He shoots in the low 90s. He still flies a single-engine Beechcraft between 40 and 60 hours a month. And every time he sees an

F-15 knife through the air over the base, he wishes he were flying it.

He never got much credit or acclaim for his pioneering work on the laser-guided bomb. The Air Force bigwigs at Wright Patterson, he says, "just wanted to keep the publicity up there, and I didn't fight them." When he retired, he received the Legion of Merit medal for his work on the smart bomb -- a lesser commendation than his commander nominated him for. When the Persian Gulf War rolled around in 1991 and Davis saw the full implications of his innovation, he was awed -- and a little bit frightened.

"Here's the thing you got to worry about -- that laser bomb goes in and kills a family or two, and that guy finds out who came up with this laser bomb, 'Oh, he's down there in Florida,' " Davis says. "And he comes down here looking for me, and puts a bullet right through my head."

So Davis kept his mouth shut all those years, despite the fact that precision airstrikes -- by shortening wars and limiting collateral damage -- have arguably saved thousands of lives, of civilians and soldiers alike.

When operational testing began in the summer of 1966, the first TI prototype missed the target by 148 feet. The second missed by 78 feet, more than twice the 30-foot accuracy TI had promised. It turned out there were glitches in the sensor, and the fins, and the electronic system. Once they were all corrected, the bomb's accuracy improved dramatically. By the fifth drop, it fell within 12 feet of the target. By the sixth, it missed by just 10 feet.

TI's competition -- North American -- wasn't doing nearly as well. Its prototype used a far more sophisticated guidance system, but experienced a string of failures in its first four drops. And it was so much more expensive than the relatively simple TI kit that North American couldn't afford to build more prototypes.

By May 1967, TI was awarded a $1.35 million contract to build 50 more bomb kits for further testing. Beyond its own success on the test range, the program was being propelled by heavy aircraft losses in Vietnam. As he and other engineers worked on perfecting the bomb, Word remembers, he saw a photograph of the Thanh Hoa Bridge standing intact amid hundreds of bomb craters on either side of the river.

In the spring of 1968, Davis and a group of test pilots and engineers from Eglin flew to Thailand to test the bombs in actual combat. The first drop on a North Vietnamese bridge was a total failure -- the bombs missed entirely.

Back then, two aircraft were needed to drop a single laser-guided bomb. A weapons officer in the back seat of one F-4 Phantom used a hand-held laser to illuminate a target, much as Davis had aimed his movie camera at the dam. And a second F-4 Phantom dropped the laser-guided bomb on a glide path that enabled its sensor to pick up the laser spot. An extensive debriefing of both pilots determined that one of them had put the laser beam on one bridge, and the other had launched the bomb at another bridge 3,000 feet away.

Once the pilots agreed on the target, the bombs started performing spectacularly. They were used briefly against bridges and other high-value targets before President Lyndon Johnson's 1968 moratorium on bombing North Vietnam. After that, they were used in Laos against small bridges and cave entrances. By the time the moratorium was lifted in 1972 and the Thanh Hoa Bridge went down, Davis's inspiration had officially become the Paveway laser-guided bomb, with a new and more accurate version featuring an enhanced laser sensor and pop-out fins. Meanwhile, advances allowed a single jet to both illuminate a target and drop a bomb.

By the end of the war in 1975, the United States had dropped more than 28,000 Paveways in Southeast Asia. The laser-guided weapons represented less than 1 percent of the 3.3 million bombs dropped during the war. But they had proved themselves on the battlefield with stunning accuracy.

The American public, sick of the endless conflict, took little notice. But the success of the laser-guided bombs and early versions of other smart bombs guided by video cameras and infrared sensors made a substantial impression throughout the U.S. military, triggering programs to perfect all three.

Says former Air Force historian Richard P. Hallion: "It was as revolutionary a development in military air power terms as, say, the jet engine."

William J. Perry understood the implications as well as anyone. A former engineering professor with a PhD in mathematics, Perry became undersecretary of defense for research and engineering -- the Pentagon's No. 3 official -- in 1977 at the beginning of the Carter administration.

It was the frigid depths of the Cold War, and the only hedge against the Soviet Union's huge advantage in conventional arms was the American nuclear capability. This presented the uncomfortable prospect of a Soviet push into Western Europe leaving the West with only two ugly options: losing, or nuclear war.

Now Perry imagined a third option: building on superior American technology, particularly in computing, to create an entirely new kind of conventional warfare. He remembers discussing this with his boss, Defense Secretary Harold Brown, himself a nuclear physicist, within their first six months in office. Together, they decided that the Soviet advantage in conventional weapons could best be countered by precision munitions, American computing prowess and a supersecret new technology called "stealth" that could make U.S. aircraft nearly invisible to Soviet air defenses.

Still licking its wounds after Vietnam, the Pentagon was quietly rearming by the time Perry arrived. Perry took it upon himself to nurture the high-tech components, despite fierce resistance from a group of critics who formed what was known at the time as the "defense reform movement." They argued that the Pentagon had become obsessed with high-tech weaponry that was wildly expensive, unproven and, most likely, ineffective on the battlefield.

Perry held his ground -- and committed huge amounts of money in a secret, "black" portion of the defense budget to building the F-117 stealth fighter. "Stealth in particular was on an incredible fast track," says Perry, now 75, a slight man, his hair thinning and gray. He notes with obvious satisfaction that the supersecret plane was developed in just four years.

Both the F-117 and the stealth B-2 bomber, also begun under Perry, feature boomerang-like designs that deflect and disperse radar waves, instead of reflecting them back at their transmitters. The aircraft are also made of materials like graphite that absorb radar waves and reduce the "radar cross-section" -- the extent to which a plane is visible on radar -- to that of a bird.

The world wouldn't see these weapons in action until the opening days of the Gulf War. But Perry foresaw their potential a decade in advance, predicting confidently in a March 1980 interview that the new precision weaponry "has to change the face of battle."

Down at Eglin, Donald Lamberson followed in Davis's footsteps as deputy commander of the Air Force's Armament Division, trying to sell the latest smart bomb technology to the rest of the Air Force. Despite Perry's support and the success of the laser-guided bomb in Vietnam, Lamberson found the going slow.

"We were flying wonderful airplanes, and we were dropping World War II munitions out of them," says Lamberson, now 71.

The fighter pilots who ran the Tactical Air Command, he says, were focused on air-to-air combat and still somewhat leery of the risks fighter jocks had to take at low altitudes dropping laser-guided bombs. And the long-range bomber pilots who ran the Strategic Air Command, he says, had even less interest in precision. One thing, and one thing only, he says, finally convinced the Air Force, if not the rest of the world, that a new way of war was at hand: Operation Desert Storm.

In the early-morning darkness, Apache helicopter gunships hugging the earth fired laser-guided Hellfire missiles and Hydra rockets at two Iraqi early-warning radars. It was January 17, 1991. The Gulf War had begun.

Minutes later, F-117 stealth fighters began attacking hardened air defense sites and command facilities in Baghdad with precision weapons, undetected by Iraqi air defenses that were seven times as dense as those around Hanoi when the United States first started using laser-guided bombs in Vietnam.

"By the time dawn broke the morning of January 17," Hallion writes in his history of the war, Storm Over Iraq, "Iraq was well on the way to losing the war . . ."

America was using its new smart bombs in a wholly new way, not just dropping large numbers of them on hard-to-topple bridges, but using them to create what Hallion calls "a simultaneous takedown, if you will, of the country."

In 1944, when 47 B-29s raided the Yawata steel works in Japan during World War II, Hallion says, "only one plane actually hit the target area, and with only one of its bombs." On the opening night of the Gulf War, by comparison, a single F-117 carrying two laser-guided bombs produced twice the destructive effect of that entire fleet.

Only about 9 percent of the munitions dropped in the Gulf War -- 7,400 of 84,200 tons -- were precision-guided, largely because stockpiles were limited. But that 9 percent was responsible for 75 percent of the damage done to strategic targets. "It was clear to everybody that those were the munitions that counted," says retired Gen. Merrill "Tony" McPeak, Air Force chief of staff at the time.

The air war lasted 37 days and quickly progressed from strategic targets around Baghdad to Iraq's troops and armor throughout the country. F-111F fighter-bombers proved particularly effective at destroying tanks with laser-guided bombs. In the cold desert night, heat from the tanks' engines shone brightly in new infrared targeting sensors used by the jets. When the air campaign had finally done all it could, the ground war lasted four days.

Shortly after the war, McPeak said that Operation Desert Storm represented "the first time in history that a field army has been defeated by air power."

Army and Marine commanders take strong exception, arguing that ground forces -- not air power -- defeated the Iraqi army and Republican Guard. Indeed, Daryl G. Press, a professor at Dartmouth College, argues in a recent paper that air power "reduced and disrupted Iraq's military," but "did not neutralize the Iraqi forces."

McPeak dismisses such arguments by Press and others, and goes even further now. "It's ever more obvious as air power begins to keep the promise that it's had all along, that it is the main action, and other action is ancillary and supporting," he says. "The facts are what they are."

Yet even in Desert Storm, air power had clear limitations that would drive war planners, bureaucrats, engineers and scientists on a search for new weapons and improved battlefield information throughout the 1990s. Air Force and Navy strike aircraft failed to destroy a single Iraqi Scud missile launcher despite enormous effort: The United States simply couldn't zero in on the moving targets in time.

The other big problem: America's new precision way of war was rendered impotent by bad weather. Saddam Hussein's air force, grounded throughout the war, couldn't stop American jets from dropping laser-guided bombs, but the clouds could.

In fact, there was already a technology that could steer bombs right through heavy cloud cover -- it just had been ignored.

Louis R. Cerrato started working at Eglin Air Force Base a quarter-century ago in an office called Development Plans. It was a lot like the unit Joe Davis headed a decade earlier, designed to find new guidance technologies for bombs. "You throw out a lot of ideas, and every once in a while something would stick," says Cerrato, now 58, a rumpled man with a bushy mustache, bald pate, white hair and wire-rim glasses.

Unlike Davis, Cerrato is an engineer with a PhD. So in 1985, when Eglin's chief scientist forwarded a study over to Development Plans on inertially guided bombs -- bombs that would use the forces of gravity and acceleration to more or less feel their way to a target -- Cerrato was intrigued. While lacking the pinpoint accuracy of laser-guided bombs, a bomb with its own inertial guidance system could land within 30 meters of a target. For a 2,000-pound bomb with a blast radius of 250 meters, that's plenty close enough to inflict heavy damage on structures and human beings.

Cerrato envisioned a bomb whose computer would be programmed with the coordinates of the target -- its destination. The bomb's computer would also receive, immediately before launch, an update on the strike aircraft's exact coordinates, its starting point. Once the bomb was dropped, gyroscopes would measure the angular movements of the weapon, accelerometers would measure increases or decreases in speed, and a computer would integrate all the data. Thus, the bomb would know where it was at all times and where it wanted to go, and the computer could calculate midcourse corrections to steer the bomb into the target.

Cerrato believed that kits could be manufactured and attached to regular dumb bombs, much like the laser bomb kits, only at lower cost. But to Cerrato, there were more important advantages of an inertially guided bomb: It would work in all weather. And it would enable a single aircraft to bomb multiple targets -- with precision, since each bomb onboard would be programmed in advance with the coordinates of the target it was supposed to strike.

But like many visionaries, Cerrato was ahead of his time. He soon discovered the same indifference that Lamberson had encountered. His bomb languished with no money to develop it further.

But as soon as the Gulf War ended in the spring of 1991, McPeak dashed off a handwritten memo to the smart-bomb guys at Eglin: "We need all-weather precision-guided munitions." And he needed them cheap, to avoid the quick exhaustion of inventory experienced in Iraq.

Thanks to Cerrato, it didn't take long. A new bomb was born, revolutionary in how it worked and how much it would cost. Called the Joint Direct Attack Munition (JDAM) and pronounced "jay-dam" in Pentagonese, it was destined to become the wildly popular Ford Mustang of smart bombs.

Terry Little, a shoot-from-the-hip type brought in to manage the program, remembers briefing McPeak on Cerrato's all-weather bomb concept, having developed a rough cost estimate of $60,000 per copy. When McPeak asked him what his cost goal was, Little -- eager to please the boss -- lopped a third off the top and said $40,000. "I just made that up," he now admits with a chuckle. "And McPeak said, 'That's not your goal, that's your ceiling.' And that was a watershed moment for the program."

Another boost came from merging technologies. Ever since 1957, when the Soviet Union launched Sputnik, the first man-made satellite, American military planners had recognized that satellites could become a high-tech aid to navigation. In 1973, a Pentagon brain trust had spent Labor Day weekend in the office devising a blueprint for the ultimate navigational aid: the Navstar Global Positioning System, known as GPS.

The system consists of 24 satellites orbiting Earth 12,660 miles in space. The satellites send radio signals to GPS receivers, which calculate how long it takes the signals to arrive, and thus how far away the satellites are. With 24 satellites in orbit, at least four, and usually eight, are above the horizon and within range of a receiver at all times. Each receiver needs signals from four satellites to calculate its longitude, latitude and altitude.

Considered the most important innovation in military command and control since the telegraph, GPS had come of age on the battlefield during the Gulf War, guiding air-launched cruise missiles and enabling tank units and troops with hand-held receivers to know precisely where they were in the desert.

Cerrato and his colleagues at Eglin saw an opportunity to correct one of the inertially guided bomb's biggest weaknesses -- the susceptibility to drift that made the inertial navigation system far less accurate than laser-guided bombs. With the new JDAM, the onboard inertial system would still guide the bomb, but it would include a GPS receiver that got updates once a second to correct for drift. The bomb's onboard computer, powered by the same chip used in an Apple computer, integrated the GPS data with the data from the guidance system to improve the bomb's accuracy to 13 meters or less.

Final production cost for the JDAM was under $20,000, thanks to the use of commercial purchasing practices and off-the-shelf components -- the company that manufactures the guidance fins also makes parts for Toro lawn mowers.

Fighting with JDAMs would still be the budgetary equivalent of dropping Toyota Camrys on enemy forces. But compared with million-dollar cruise missiles and the latest laser-guided bombs, which cost about $60,000, the JDAM was, for the Pentagon at least, an incredible bargain.

In 1999, during NATO's 78-day air war in Serbia, when bad weather prevented the use of laser-guided bombs against the Zezeljev Bridge over the Danube at Novi Sad, the bridge was dropped with a barrage of JDAMs.

The attack was eye-opening. The new bombs did not have the precision of laser-guided weapons. But they were accurate enough -- and they are, in fact, getting better. The JDAM's accuracy is now down from 13 meters to a range of four to six meters, with the accuracy of GPS itself improving.

But even in victory, the air campaign revealed ongoing weaknesses. B-2 bombers severely damaged the Chinese Embassy in Belgrade with five JDAMs, which hit their target with exquisite accuracy. The only problem was the intelligence was wrong -- U.S. war planners thought they were targeting a Yugoslav arms agency and hit the embassy by mistake.

The disastrous attack, which killed three employees in the embassy and seriously damaged U.S.-Chinese relations, showed how dependent precision air attack was on good intelligence -- and how faulty U.S. intelligence could be.

High on the battlefield in Afghanistan, a Northern Alliance commander watched Taliban forces moving in a valley below. To advance, he needed the enemy troops and vehicles hit from the air within 24 hours. It was November 2001, a month of nonstop U.S. airstrikes into the war on terrorism.

U.S. Special Forces troops, inserted into the battlefield to designate targets for Northern Alliance forces, relayed the commander's request via satellite to the U.S. command center in Saudi Arabia. The command center radioed an incoming B-52 bomber and put its pilots in contact with the Special Forces troops below. Those soldiers pointed a laser designator at the Taliban troops to fix where they were located. Then they plugged the designator into a portable GPS receiver to derive their GPS coordinates and relayed the coordinates back to the bomber by radio.

JDAMs started falling on the Taliban within 20 minutes of the Northern Alliance commander's request. The combination of the Special Forces spotters on the ground and the new bomb in the air enabled the lumbering B-52, an aircraft designed in the 1950s, to fly what has traditionally been considered one of the most demanding battle missions -- providing "close air support" of ground troops.

U.S. pilots have dropped more than 5,000 JDAMs in Afghanistan, about half the initial inventory. Use of the bomb has been so extensive that worried Pentagon war planners have now told Boeing, the bomb's manufacturer, to crank up its production line. Instead of 82,000 bombs, they now want Boeing to build 230,000.

But with the JDAM and a full array of laser-guided bombs in hand, the precision revolution's new frontier in Afghanistan was all about information: fusing data from multiple sensors and moving it almost instantaneously to

pilots waiting to strike. The most sophisticated "sensors" of all were human beings on the ground, Special Forces teams capable not only of distinguishing an SUV from a school bus, but knowing who was inside each.

Yet even as American spotters on the ground were proving their worth, the Afghan air war suggested a future when human beings on the battlefield would be largely unnecessary. Predator reconnaissance drones -- remote-controlled aircraft equipped with video, infrared and radar sensors and armed with laser-guided Hellfire missiles -- sighted enemy targets and fired at them. Sensor and shooter became one and the same.

The Predator can linger for as long as 20 hours over a battlefield at up to 25,000 feet, beaming information back to commanders on the ground -- or directly to other aircraft. Ground operators on the other side of the planet fly the aircraft by remote control, via satellite uplink: Instead of looking out the cockpit window, a Predator "pilot" and two sensor operators look at streaming video beamed back from the aircraft. The Predator was joined in Afghanistan by the high-altitude Global Hawk drone, which can fly higher (65,000 feet) and longer (24 hours).

At the start of the war on terrorism, Gen. John P. Jumper, the Air Force chief of staff, ordered a research team to investigate the possibility of feeding Predator targeting video directly to AC-130 Spectre gunships. One of the most fearsome weapons against mobile targets in the Air Force inventory, the AC-130 is equipped with three side-mounted cannons, including a 25mm Gatling gun that shoots 1,800 rounds a minute. Two months after Jumper's directive, AC-130 crews were watching Predator video of their targets before they flew within firing range.

Last November, for example, commanders at the air operations center in Saudi Arabia got intelligence about an impending meeting of Taliban and al Qaeda leaders. They dispatched a Predator to the general vicinity of where the meeting was supposed to take place. Soon, a radar aircraft detected a convoy moving on the ground. Its data was fed to the Predator, which began surveilling the convoy with its video sensor. Commanders hundreds of miles away viewed footage of the convoy rolling toward a small mosque. They called in an AC-130 to attack. As it approached, video from the Predator was fed directly to the gunship so that its crew could locate their targets even before they flew over. When the plane arrived over the meeting place near the mosque, it opened fire immediately, shooting individuals as they tried to flee to waiting vehicles. The mosque, designated as a "no strike" target, remained untouched.

War planners at the Pentagon like to contrast this with the situation that existed in the Gulf War, when messenger aircraft had to fly the next day's air war master plan out to the Navy's aircraft carriers on computer diskettes. The targets listed were, on average, 72 hours old. In Afghanistan, by contrast, many targets -- instead of being identified 72 hours earlier -- have emerged only after the planes took off from the aircraft carrier. The Navy reports a 65 percent success rate in attacking those "emerging" targets, some of which were actually moving at the time they were hit.

Last month in Yemen, a machine did it all. A lone Predator, operated from a distant base, picked up a suspected al Qaeda target in a sport utility vehicle, tracked the SUV as it sped along a highway, then destroyed it with a Hellfire missile, killing six suspected terrorists.

The next war will be even higher tech. If it is against Iraq, which has considerably more sophisticated air defenses than Afghanistan, the U.S. military will undoubtedly employ new generations of so-called "stand-off" weapons it didn't need to use in Afghanistan. Stand-off weapons enable the U.S. military to launch precision-guided munitions from well beyond the reach of a more capable adversary's air defenses.

The most impressive is the newly developed, stealthy Joint Air-to-Surface Stand-Off Missile, an air-launched cruise missile that has pinpoint accuracy from 200 miles. It is guided by a GPS-aided inertial system until late in its flight, when an infrared seeker programmed to recognize specific features of a target is activated. The seeker then guides the bomb into the target with three-meter accuracy, so it can be used, for example, to recognize and destroy the third span of a bridge or hit the fourth floor of an office building.

There is also the new Joint Stand-Off Weapon, a GPS-guided glide bomb that can find a target from 15 to 47 miles away, depending upon speed and altitude. (A JDAM's range is only six to 13 miles.)

And there is the Sensor Fuzed Weapon, designed to pulverize tanks, light armor and troops over a vast 30-acre landscape. The weapon is dropped from an airplane, opens in midair and releases 10 projectiles, which float down to the battlefield on tiny parachutes. When the projectiles are close to the ground, rocket motors spin them back up, where each of the 10 releases four warheads the size of soda cans, 40 in all. Each warhead is equipped with infrared and laser sensors. The infrared sensor scans the battlefield looking for preprogrammed heat signatures, like those emitted by tanks and other vehicles. Once a target is located, the laser sensor then guides the warhead into it. If the infrared sensor locates a tank or other vehicle, it fires an armor-piercing slug. But if no vehicles are spotted, the sensor detonates the warhead just above the ground to spray the battlefield with thousands of lethal fragments, or steel rain.

In addition to these new bombs, the bomb scientists at Eglin and other Air Force research installations have produced an upgraded bunker-busting warhead that features an elongated spike made of a nickel-cobalt steel alloy that enables the warhead to puncture 11 feet of reinforced concrete, the equivalent of more than 100 feet of soil. A "smart fuse" for such penetrating warheads uses an accelerometer to count the number of floors penetrated before detonation.

These advances, however, pale in comparison with the technology now under development and expected to reach maturity in a decade, at most.

What may be the smartest bomb of all is now in development at Eglin. The winged bomb would find targets, track them, and then decide what type of warhead would be best to destroy them -- all on its own.

Powered by a miniature turbo jet engine and weighing 85 pounds, each of the bombs -- dubbed Low Cost Autonomous Attack Systems -- would fly 55 miles to a target and then loiter over the battlefield for 15 minutes, searching a 25 square-kilometer area with a sensor that uses a laser beam to create images of targets moving on the battlefield.

Identifying a target involves something called "automatic-target-recognition algorithms" that enable the sensor to digitally compare what it is seeing on the ground with models already programmed into its onboard computer. "It can distinguish an SA-6 radar from an SA-8 telar," says James Moore, project manager. "Or an SA-6 radar from an SA-6 launcher, each having the same chassis. That's pretty good. Once it's convinced itself that it's got the right target, it would then start diving down on it."

Here, the bomb would exhibit its final act of autonomous intelligence, arming the warhead, either as an armor-piercing slug or as antipersonnel fragments, depending upon the type of target it has fixed. The cost of each of these "brilliant" bombs is only $63,000, roughly the same as the latest model laser-guided bombs.

The notion of a brilliant bomb loosed over a battlefield striking targets of its own computerized volition strikes some as the ultimate in precision air warfare, and others as the ultimate nightmare.

Indeed, the future implications of America's ever-increasing precision airstrike prowess are hard to fully fathom. One unsettling possibility: Anyone can access GPS signals by purchasing an off-the-shelf receiver, and yes, there is concern that terrorists could launch homemade GPS-guided cruise missiles from ships off the coast of the United States. Another: that America's unmatchable military technology will simply drive terrorist and rogue states toward more accessible, and more destructive, responses, including chemical, biological and nuclear weapons, or airliners full of innocent civilians.

But air power enthusiasts epitomized by John A. Warden III, a retired Air Force colonel who played a leading role in devising the Gulf War air campaign, think America's growing global strike prowess is the solution: He argues that even rogue nations will have to think twice about aggressively building weapons of mass destruction given America's ability to attack without warning and strike with extreme precision.

In a new book called The Precision Revolution, Michael Russell Rip and James M. Hasik note the danger of forcing rogue nations off the conventional battlefield, but hasten to add that "this power to achieve almost nuclear results with ordinary explosives could lead to a new degree of conventional deterrence. Aggression could conceivably be thwarted by merely revealing American forces' target lists . . . or the photographs of a foreign leader snapped by a Global Hawk."

Warden argues that air power superiority is central to national security. "I couldn't be more encouraged," he says. "As long as we keep working this stuff aggressively, there's no reason why we can't maintain this dominant military position into the indefinite future."

Warden is, however, far from satisfied. As good as U.S. air commanders now are at hitting precisely what they're aiming at, he says, they aren't thinking small enough. "If you know there are a bunch of bad guys in a room, in theory you have no reason to blow up the whole house," Warden says, arguing that precision strikes, at least theoretically, give the United States the capability to wage war "without killing anybody that we don't want to kill specifically."

Sarah Sewall, a program manager at the Carr Center for Human Rights Policy at Harvard's Kennedy School of Government, says the United States is a long way from realizing Warden's vision. But even if the U.S. military had the technical means, Sewall argues, it is nowhere near having the information needed to make smart bombs smart enough. The problem now, she says, isn't the smart bombs, it's the intelligence that decides where they're used -- a problem demonstrated most recently in a June attack on an antiaircraft battery in Afghanistan that accidentally killed at least 40 innocent civilians at a nearby wedding party. "You can precisely hit," she says, "exactly the wrong thing."

Kenneth Roth, executive director of Human Rights Watch in New York, believes the greatest danger of these new weapons may be an outgrowth of the growing confidence that they can decapitate an enemy while minimizing civilian casualties. With the presumption that war's ugly side has diminished, war planners may be too quick on the trigger. "The increased reliance on precision weapons is not a substitute for critical self-scrutiny," Roth says. "And the Pentagon has generally refused to do serious analysis of whether, in fact, it did everything [in Afghanistan, Yugoslavia, Bosnia and the Gulf War] to avoid civilian casualties."

The U.S. Central Command, for example, examined the circumstances surrounding the deaths at the June wedding party, which came during an attack by an AC-130 on the antiaircraft battery, and concluded the attack was based on accurate intelligence and within the rules of engagement, even though it killed dozens of women and children. The gunship's crew members were able to see where the antiaircraft fire was coming from but could not see that a wedding was taking place nearby.

Bill Perry, the father of stealth, takes a middle ground in the debate over the benefits of precision strike. Now teaching at Stanford, he acknowledges that America's overwhelming military superiority probably has pushed the rogue states to attempt to develop weapons of mass destruction sooner than they otherwise might have.

But he remains true to his past and optimistic about the benefits of technology, as long as final decisions about striking targets are made by humans, not machines cuing other machines to automatically open fire. "I hope we are wise enough to use it appropriately, which means keeping well-trained, thoughtful humans in the loop."

Vernon Loeb covers the Pentagon for The Post. He will be fielding questions and comments about this article at 1 p.m. Monday on