Peace on Earth by Stanislaw Lem Page A

sector of each country was installed as a self-evolving testing range. Take a look.” He held a small flat box. The different sectors of the moon lit up, until the globe was as bright as a Chinese lantern. “The larger areas belong to the superpowers. Of course we know what we put there: the Agency acted as transporter, after all. We also dug the foundations for the simulators. Each sector has two simulators surrounded by a production compound. The sectors cannot fight each other; it’s impossible. One simulator designs new weapons and the other works to counter them. Both are computers programmed on the sword-and-shield principle. It’s as if each nation put on the moon a computer that played chess with itself. Except that the game is played with weapons instead of chess pieces and everything can change: the moves, the pieces, the board, everything.”
    “You mean,” I asked, surprised, “there’s nothing up there but computers simulating an arms race? Then how is this a threat to Earth? Surely the simulation of a weapon no more dangerous than a piece of paper…”
    “Oh no! The weapons that survive selection go into real production. The whole problem is when. You see, the simulators design not just a new weapon but a whole new system of warfare. These are, of course, nonhuman systems. The soldier becomes one with the weapon. Think of natural evolution, the struggle for existence, Darwin. One simulator designs, say, a kind of predator, and the other finds its weakness in order to destroy it. Then the first simulator thinks up something new, and the second parries that too. In principle a contest like this, with endless improvements, could go on for a million years—but each sector after a certain time must begin actual production of the weapons. The time—and the effectiveness required of the prototypes—was determined in advance by the programmers of each nation. Because each nation wanted to have an arsenal of real weapons on the moon, not just simulations like blueprints on paper. Therein lies the rub, the contradiction. Do you understand?”
    “Not entirely. What contradiction?”
    “Simulated evolution proceeds much more swiftly than natural evolution. He who waits longer obtains the better weapon. But for as long as he waits, he is defenseless. While the one who accepts a shorter simulation run will obtain his weapon sooner. We call this the coefficient of risk. Every nation, placing its military might on the moon, had to decide first whether it wanted better weapons later or poorer weapons sooner.”
    “Curious,” I said. “And what happens when sooner or later production begins? The weapons are stockpiled?”
    “Some of them, perhaps. But only some. Because then an actual, not a simulated, battle begins—of course only within the given sector.”
    “Something like maneuvers?”
    “No. On maneuvers the fighting is an exercise, soldiers don’t die, whereas there”—the director pointed to the colorful moon—“genuine combat is taking place. But always within the borders of each sector. Neighbor cannot attack neighbor…”
    “So the weapons attack and destroy each other in the computer first, and then for real. And what then?”
    “Good question! No one knows. There are basically two possibilities. Either the arms race reaches a limit or it doesn’t. If it does, this means ‘an ultimate weapon’ exists and that the simulated arms race has finally arrived at it. The weapon cannot defeat itself, and thus ensues a state of permanent equilibrium. An end to progress. The lunar arsenals fill with that weapon, which has passed the final test, and nothing more happens. This is what we would like.”
    “But it isn’t so?”
    “Most likely not. In the first place, natural evolution has no end. It hasn’t because no ‘ultimate’ organism exists, that is, one which is perfect in survivability. Every species contains a weakness. Secondly, the evolution on the moon is artificial, not natural. And