A vision of dozens or hundreds of nuclear-armed missiles on ready standby to deal with threatening asteroids or comets has been offered. However premature in this particular application, it seems very familiar; only the enemy has been changed. It also seems very dangerous.

The problem, Steven Ostro of JPL and I have suggested, is that if you can reliably deflect a threatening worldlet so it does not collide with the Earth, you can also reliably deflect a harmless worldlet so it does collide with the Earth. Suppose you had a full inventory, with orbits, of the estimated 300,000 near-Earth asteroids larger than 100 meters—each of them large enough, on impacting the Earth, to have serious consequences. Then, it turns out, you also have a list of huge numbers of inoffensive asteroids whose orbits could be altered with nuclear warheads so they quickly collide with the Earth.

Suppose we restrict our attention to the 2,000 or so near-Earth asteroids that are a kilometer across or bigger—that is, the ones most likely to cause a global catastrophe. Today, with only about 100 of these objects catalogued, it would take about a century to catch one when it’s easily deflectable to Earth and alter its orbit. We think we’ve found one, an as-yet-unnamed[33] asteroid so far denoted only as 1991OA. In 2070, this world, about 1 kilometer in diameter, will come within 4.5 million kilometers of the Earth’s orbit—only fifteen times the distance to the Moon. To deflect 1991OA so it hits the Earth, only about 60 megatons of TNT equivalent needs to be exploded in the right way—the equivalent of a small number of currently available nuclear warheads.

Now imagine a time, a few decades hence, when all such near-Earth asteroids are inventoried and their orbits compiled. Then, as Alan Harris of JPL, Greg Canavan of the Los Alamos National Laboratory, Ostro, and I have shown, it might take only a year to select a suitable object, alter its orbit, and send it crashing into the Earth with cataclysmic effect.

The technology required—large optical telescopes, sensitive detectors, rocket propulsion systems able to lift a few tons of payload and make precise rendezvous in nearby space, and thermonuclear weapons—all exist today. Improvements in all but perhaps the last can be confidently expected. If we’re not careful, many nations may have these capabilities in the next few decades. What kind of world will we then have made?

We have a tendency to minimize the dangers of new technologies. A year before the Chernobyl disaster, a Soviet nuclear power industry deputy minister was asked about the safety of Soviet reactors, and chose Chernobyl as a particularly safe site. The average waiting time to disaster, he confidently estimated, was a hundred thousand years. Less than a year later… devastation. Similar reassurances were provided by NASA contractors the year before the Challenger disaster: You would have to wait ten thousand years, they estimated, for a catastrophic failure of the shuttle. One year later… heartbreak.

Chlorofluorocarbons (CFCs) were developed specifically as a completely safe refrigerant—to replace ammonia and other refrigerants that, on leaking out, had caused illness and some deaths. Chemically inert, nontoxic (in ordinary concentrations), odorless, tasteless, non-allergenic, nonflammable, CFCs represent a brilliant technical solution to a well-defined practical problem. They found uses in many other industries besides refrigeration and air conditioning. But, as I described above, the chemists who developed CFCs overlooked one essential fact—that the molecules’ very inertness guarantees that they are circulated to stratospheric altitudes and there cracked open by sunlight, releasing chlorine atoms which then attack the protective ozone layer. Due to the work of a few scientists, the dangers may have been recognized and averted in time. We humans have now almost stopped producing CFCs. We won’t actually know if we’ve avoided real harm for about a century; that’s how long it takes for all the CFC damage to be completed. Like the ancient Camarinans, we make mistakes.[34] Not only do we often ignore the warnings of the Oracles; characteristically we do not even consult them.

The notion of moving asteroids into Earth orbit has proved attractive to some space scientists and long-range planners. They foresee mining the minerals and precious metals of these worlds or providing resources for the construction of space infrastructure without having to fight the Earth’s gravity to get them up there. Articles have been published on how to accomplish this end and what the benefits will be. In modern discussions, the asteroid is inserted into orbit around the Earth by first making it pass through and be braked by the Earth’s atmosphere, a maneuver with very little margin for error. For the near future we can, I think, recognize this whole endeavor as unusually dangerous and foolhardy, especially for metal worldlets larger than tens of meters across. This is the one activity where errors in navigation or propulsion or mission design can have the most sweeping and catastrophic consequences.

The foregoing are examples of inadvertence. But there’s another kind of peril: We are sometimes told that this or that invention would of course not be misused. No sane person would be so reckless. This is the “only a madman” argument. Whenever I hear it (and it’s often trotted out in such debates), I remind myself that madmen really exist. Sometimes they achieve the highest levels of political power in modern industrial nations. This is the century of Hitler and Stalin, tyrants who posed the gravest dangers not just to the rest of the human family, but to their own people as well. In the winter and spring of 1945, Hitler ordered Germany to be destroyed—even “what the people need for elementary survival”—because the surviving Germans had “betrayed” him, and at any rate were “inferior” to those who had already died. If Hitler had had nuclear weapons, the threat of a counterstrike by Allied nuclear weapons, had there been any, is unlikely to have dissuaded him. It might have encouraged him.

Can we humans be trusted with civilization-threatening technologies? If the chance is almost one in a thousand that much of the human population will be killed by an impact in the next century, isn’t it more likely that asteroid deflection technology will get into the wrong hands in another century—some misanthropic sociopath like a Hitler or a Stalin eager to kill everybody, a megalomaniac lusting after “greatness” and “glory,” a victim of ethnic violence bent on revenge, someone in the grip of unusually severe testosterone poisoning, some religious fanatic hastening the Day of judgment, or just technicians incompetent or insufficiently vigilant in handling the controls and safeguards? Such people exist. The risks seem far worse than the benefits, the cure worse than the disease. The cloud of near-Earth asteroids through which the Earth plows may constitute a modern Camarine marsh.

It’s easy to think that all of this must be very unlikely, mere anxious fantasy. Surely sober heads would prevail. Think of how many people would be involved in preparing and launching warheads, in space navigation, in detonating warheads, in checking what orbital perturbation each nuclear explosion has made, in herding the asteroid so it is on an impact trajectory with Earth, and so on. Isn’t it noteworthy that although Hitler gave orders for the retreating Nazi troops to burn Paris and to lay waste to Germany itself, his orders were not carried out? Surely someone essential to the success of the deflection mission will recognize the danger. Even assurances that the project is designed to destroy some vile enemy nation would probably be disbelieved, because the effects of collision are planet-wide (and anyway it’s very hard to make sure your asteroid excavates its monster crater in a particularly deserving nation).