This was, obviously, counter to the spirit of disarmament and reducing the world’s nuclear arsenal, which has been the purported goal of the world’s nuclear states since the 1960s. The tests weren’t about ensuring that America’s nukes still worked or learning about the fundamental physics of the weapon. They were about building bigger and better bombs. “Very few of the tests were reliability tests, where you blow it up to see if it still works,” says Gusterson. “They were almost all tests to develop new designs.”

The US ended all underground nuclear tests in the early 1990s in the lead-up to the Comprehensive Nuclear Test Ban Treaty, despite protests from the heads of the nation’s three national weapons labs—Lawrence Livermore, Sandia, and Los Alamos—who fought “tooth and nail” to prevent the ban, says Gusterson. They were concerned, he says, that a ban would reduce the reliability of America’s nukes and prevent the next generation of nuclear weapons designers and engineers from learning the tools of the trade. But perhaps most importantly, they saw the ban as a threat to the labs’ very existence. All three had been founded to further the development of America’s nuclear arsenal. What was the point of keeping them around if not to blow up their creations?

Mark Chadwick, the chief scientist in the Los Alamos Weapons Physics Directorate, arrived at the national lab in 1990 fresh out of a physics doctoral program at Oxford. At the time, he says, there was a lot of debate among the Los Alamos scientists about the future of the lab, or whether it would have a future at all. “Some thought the labs would really end up struggling to find business and that the nuclear deterrence mission would sort of fade away,” Chadwick recalls. “Overall, the pessimism that the national security mission wouldn’t remain important proved wrong. And fairly quickly, in fact.”

The US conducted its last explosive nuclear test in September, 1992. Today, the nation’s nuclear weapons research is focused on reliability testing and maintenance of the roughly 4,000 active warheads in its arsenal, a program broadly referred to as “stockpile stewardship.” After the test ban, the US government lavished funding on the new stewardship program to keep the nation’s weapons up to snuff. The so-called virtualization of US nuclear tests meant that weapons scientists would employ the most powerful lasers and supercomputers in the world to understand these weapons instead of blowing them up. Physicists at the labs work on the best experimental equipment that money can buy, and their funding has ballooned under the Trump administration. “Business is booming, even without nuclear testing,” says Gusterson.

At the heart of the US stockpile stewardship program is Lawrence Livermore National Laboratory, a sprawling complex across the bay from San Francisco. It’s home to the National Ignition Facility, which uses the most powerful laser in the world to re-create the conditions found in the heart of an exploding nuclear bomb. “It’s not so much that it replaces nuclear testing, but it’s a very different, richer perspective on what’s happening in an operating weapon,” says Kim Budil, the director of the lab.

Nuclear tests have always served a variety of purposes. Their primary one, of course, has been deterrence—an ever-increasing show of strength meant to discourage America’s allies from ever hitting the big red button. But even back when the military detonated live nukes, its architects were doing everything they could to figure out exactly what was happening inside. Each bomb was outfitted with tens of millions of dollars worth of sensors designed to capture data in the fraction of a fraction of a second before they were destroyed. Virtualization now allows scientists to dig deeper into the physics of the bomb.