For almost 20 years, he rode the cutting edge of weapons design, all the while telling himself that it was good for the United States to have a powerful arsenal. The better he could make the weapons, Taylor reasoned, the less likely it was that anyone would be insane enough to actually use them.
But in 1964 Taylor took a job at the Pentagon providing technical advice on the real-world effects of nuclear bombs. Suddenly, instead of theorizing about how to make a bomb smaller, or lighter, he was on the military side, calculating how much destruction different weapons might rain on our enemies. He drew circles on maps of Moscow, to see which of our bombs could wipe out the largest chunk of the city.
He lost his taste for it.
"It was finding out, which I had not known before, how many nukes we had at the time," Taylor recalls. "I thought at most a couple thousand. In fact we had about 35,000 at that time. That's subsequently been declassified, so I can say it.
"I was dumbfounded. I saw no possible reason we should have so many."
Taylor switched sides, and has since devoted his life to lobbying for disarmament and fighting the spread of nuclear technology.
To Taylor, the government's latest foray into nuclear weapons research, the Stockpile Stewardship program, is a disturbing venture. Not only does it promise to sustain the country's addiction to nuclear weaponry, he believes, but it holds the potential for the U.S. to take a huge scientific leap and finally design the weapon that has long eluded our best scientific minds -- the pure fusion bomb.
Understanding Taylor's fear requires some explanation of how nuclear wea-pons work.
When a modern warhead explodes into a fireball, there are actually two nuclear reactions. The first is a fission reaction, in which the atoms in a radioactive material are split apart, unleashing energy.
Each warhead contains a charge of conventional high explosives, like TNT. These high explosives surround a small "pit" of plutonium or enriched uranium. When the explosives go off, they compress the pit, splitting the atoms and starting a fission reaction.
This first reaction, in turn, produces the heat and pressure needed to start the second reaction, a fusion reaction, in which atoms in a thermonuclear fuel like deuterium or tritium are smashed together, releasing even greater amounts of energy.
A fission reaction in and of itself is a powerful beast, capable of producing a respectable fireball. The two bombs dropped on Japan during World War II were both simple fission devices.
But even bigger firepower comes from a fusion reaction, and our arsenal has been significantly more fearsome since U.S. scientists developed the two-step bomb.
No one, however, has yet been able to bypass the first step by figuring out how to trigger a fusion reaction without using a fission reaction to jump-start it. That scientific hurdle has been one of the greatest impediments to anyone -- small nations, terrorist groups, unstable dictators, or whomever -- who might be trying to develop his own nuclear weapons.
Simply put, you need a fission reaction to spark fusion. But to achieve fission, you need either plutonium or enriched uranium, two materials that are very hard to come by. As long as the U.S. and other nuclear nations can keep a tight leash on supplies of these materials, it will be exceedingly difficult for anyone else to make warheads.
The fuels needed for fusion reactions, on the other hand, are plentiful and common elements like lithium and hydrogen. If someone can find a way to spark the fusion reaction without using plutonium or enriched uranium, nuclear weapons will suddenly become much easier to build, and within the scientific grasp of many more would-be nuclear powers.
And that, almost precisely, is what the National Ignition Facility -- the massive new laser being built at Lawrence Livermore National Laboratory -- is designed to do. The NIF, as it is called, will subject tiny pellets of lithium, deuterium, or other fuels to intense pressure and heat, trying to start a small fusion reaction.
"Its function is, for the first time in the history of human beings, to make nuclear explosions without using any fissile material [plutonium or enriched uranium]," Taylor says. "We will be crossing over into a whole new arena."
Many people disagree with Taylor. While his theory may sound plausible, they say, practically speaking there is no way to build a pure fusion bomb. It's a dream, a physicist's chimera, which hasn't proven possible after 50 years of nuclear weapons research.
"I've never seen any even hypothetical analysis of a pure fusion weapon that looked possible," says William Hogan, the senior scientist for the NIF at Lawrence Livermore. "As a weapon, it's not a practical one."
The design of the NIF itself, Hogan says, proves a pure fusion bomb is unlikely ever to be built.
When finished, the NIF laser will be about as big as a football field. It will have 192 beams, all aimed to converge on a single point the size of a BB. If it works, it will produce pure fusion, Hogan concedes. But how do you fit a football field-size laser into a warhead? "You can't make the laser small enough to be a deliverable weapon," he says.
But Taylor says that's a misleading argument. The amount of energy used by the NIF will not be that great, he points out. Although it will be the most powerful ever built, the laser will only use about the amount of electricity needed to heat a few cups of coffee. The reason the NIF will be so powerful is because it will take that energy and compress it into a burst lasting three-billionths of a second.
Once scientists have seen how pure fusion works, Taylor says, it may prove practical to take other sources of energy -- like high explosives -- and devise ways for them to deliver the same amount of energy in a short, concentrated burst.
"The energy handled by those lasers is the equivalent of only a few pounds of high explosives," Taylor says.
David Dearborn, a weapons designer at Livermore, thinks Taylor is way off-base. "That's such an incredibly huge assertion that it's not anything to do with science," Dearborn says. "They have tried pure fusion in the past. It's hard. If it was easily doable with high explosives, we would have done it."
Maybe they would have, if Taylor still worked in weapons. His genius has always been redefining what others believe to be possible, and he has often been right.
It's possible the NIF laser won't springboard scientists to a pure fusion bomb. But it's still probable enough that Taylor is very worried about the prospect. If the secret to a pure fusion bomb is unlocked, he says, it will only be a matter of time before that knowledge will spread, out there for anyone who wants to put it to use.
Nuclear weapons will no longer be the sole province of countries that control supplies of plutonium and enriched uranium. Anyone will be able to jump into the game.
"With all the vigor we can muster, some of us are saying the U.S. -- our country -- led us into the nuclear age, and it's time we led the world out," Taylor says. "And we're doing the opposite. And it's getting worse and worse as time goes by, as things like the NIF get closer to completion.