Europa Strike (12 page)

“Why not? They will anyway.”

“True enough. But…remember the
Heavenly Lightning
?”

“The Chinese research ship that left geosynch three weeks ago? Sure.”

“Nineteen days ago, they detonated a nuclear warhead in what they claimed was a test, but because of the plasma cloud, we lost touch with her for a period of several hours. We think it's just possible that she used a railgun to launch antimatter warheads at the
Kennedy
. On that trajectory, they would have come out of the sun, and at extremely high velocity. We estimate 300-plus kps. If the warheads were stealth designs, and if they mounted sensor and maneuvering systems so they could come in as a smart weapon with terminal track-and-control, they might have been inside the
Kennedy
's defense network before her AI even realized they were there.”

And Robbie never had a chance. Somehow, she held the tears at bay.

“It's still just a theory,” Talbot had told her. “The
Lightning
would have had to generate absolutely incredible ac celerations in some sort of onboard railgun to delivery a package at that speed.”

“What kind of accelerations?”

“Well, the
Lightning
is about 200 meters long. For a muzzle velocity of over 300 kps? Something on the order of 25 million Gs.”

“That's…impossible.”

“Agreed. With her power system, she might have managed a million Gs, for a few shots. Maybe. But…there are some other possibilities we're looking into as well.”

It was a fascinating problem in applied military science
, she'd thought. How to carry out a bombardment of a target at
extreme
range, across several astronomical units. You could fire a missile at such ranges, but your target would be certain to see it coming, and point-defense lasers, aimed by AIs with very fast reaction times, could knock down just about any solid object approaching a naval warship while it was still several thousand kilometers away.

But an artillery shell, with a course correction at the far end of the trajectory for pinpoint smart-weapon accuracy, could do it.
If
you could fire the shell at a high enough velocity.

Her mind turned the problem, desperate to stay cold and calculating and remote and not think about the reality of the target—about Robbie. Like tanks, heavy artillery in the classic battlefield sense was long extinct. As far back as the end of the twentieth century, missile launchers and heavy artillery had more and more been evolving as
mobile
artillery, with the idea of being able to launch a few rounds before enemy radar pinpointed your position and plastered you with counterbattery fire and air strikes. Really big stationary artie batteries and missile launchers were dinosaurs, easily killed, and ineffective unless massed in large, logistically complex numbers. Even a relatively small artillery shell or missile could be tracked by radar or lidar and brought down by point defense lasers. For the past eighty years, at least, the trend had been toward smaller, lighter systems—especially shoulder-launched weaponry like M-614 Wyverns, artillery support a single man could carry with him.

It appeared, however, that the Chinese might have just revived the old idea of long-range bombardment, but only by carrying the concept to mind-numbing extremes. How would you defend against such an attack? Improved sensors, perhaps…or employing picket systems or satellites to warn of high-speed incoming rounds.

Or by taking closer note of the movements of suspicious warships like the
Heavenly Lightning
, and not letting them deploy in such a way that they became a threat.

Robbie!

Blinking back the tears, she looked out the port-side window, at the purple curve of the far horizon to the north, an arc pinned beneath the black of space and the white glare off the thick clouds below. The sun was low in the west astern, casting each swirl, each bump of the clouds into sharp, three-dimensional relief. She felt a series of bumps transmitted through the deck. They were over the Great Plains now, and beginning the descent toward Washington.

She'd asked for—and General Talbot had granted—a special leave, time for her to go back and be with her family, at least until more was known. Damn…home was where she
should
have been all along.

There'd been a time, a century ago or so, when enlisted personnel were actually discouraged from marrying, when it was assumed that female Marines would get married or pregnant and have to leave the service. No more. Lots of Corps families had two or more members in the Marines nowadays; the Corps went out of its way to station family close together, though, of course, the needs of the service always came first. In her case, she'd been stationed close to home, at Quantico, until her posting as CO of 1-MSEF, which meant she had to go to V-berg. They'd been good about letting her shuttle back and forth between the coasts on weekends, but it had been a grind, a damned nightmare, and one weekend out of three she couldn't get home anyway.

Her kids needed her. She needed her kids—and Rob. Maybe it was time to think about an early retirement.

Her PAD chirped. Reaching to her holster, she pulled the device free and plugged it into the flatscreen on the back of the seat in front of her, for a bigger picture. The screen flashed, then showed the words
SECURE TRANSMISSION INCOMING
.

She glanced around her. She was on a commercial flight, but the seat next to her was vacant and none of the other passengers was paying attention. She slipped the earphone into her ear and set the sound to plug only; her conversation would be private enough. She thumbed the accept key.

It was General Talbot. “Hello, Colonel.”

“General,” she said, keeping her voice low. “I'm not alone at the moment.” She wondered if she should go to the rest room to take the call. Or ask a steward if she could use the galley/food-prep compartment.

“That's okay,” he told her. “Just listen and don't comment out loud. We have some new information about the…situation.”

“Yes, sir?”

“Word just came through a few minutes ago. Apparently, the
Franklin Delano Roosevelt
was also attacked and destroyed shortly after the
Kennedy
. We didn't know sooner, of course, because of the time delay from Jupiter space, and because the bureaucracy in Washington right now is falling all over its own feet.

“Triple-N already has the story, of course, but they haven't put it all together yet. This confirms that we have been attacked, and that it was the
Heavenly Lightning
that launched it. As of 2035 hours Zulu this evening, we are on a war footing. We expect the president will make an announcement either later tonight, or sometime tomorrow morning.”

She took another glance around. No one was listening. “What…what about our people at Europa, sir?”

“Status unknown at this time. If they were operating according to sched, the first troops should already have been on the surface. But our sensors picked up three major explosions, one of them very close to the CWS facility itself. It looks as though the Chinese have deliberately destroyed our facility on Europa, our Marine transport in orbit, and the Peaceforce cruiser that was in a position to intercept their transport. The
Star Mountain
is now scheduled to reach Europa within three days. They undoubtedly have troops on board, and intend to secure whatever is left of our installation there, and assume the task of contacting the Europan intelligence first.”

“I take it, sir, you want me back at V-berg, stat.”

He shook his head. “Not…immediately, Kaitlin. Take a few days at home. But my staff will be preparing a readiness report, and we need to know we can count on 1-MSEF. We may be looking at a relief expedition of some sort, depending on the current status of our Europan base.”

“I understand, sir.”

“Check in with my office once you get home, and be prepared for a quick ride back to V-berg.”

“Aye, aye, sir. Colonel Frickerson can have any numbers you need for your report, and you can always access Yukio.” Yukio was the persona of her secretary AI, resident in her PAD, but who also roamed the computer net at Vandenberg. “I can tell you right now that we're ready to go…I'd say on forty-eight notice.”

“That's reassuring, Colonel. Thank you. I…don't expect we'll need you back here for, oh, say, seventy-two hours yet. Just stay in touch, just in case.”

“Aye, aye, sir.”

“Kaitlin…I'm sorry about your son.”

She clenched her jaw before replying. “We don't know he's dead, sir.”

“No. And there will be a search for survivors. But as of tonight, he is being listed as ‘missing, presumed dead.' You have to know that it isn't likely that anyone survived on the
Kennedy
.”

“I…know, sir.”

“Okay. Give my best to Rob.”

“Yes, sir. I will.”

The screen went dark before coming back up with her PAD's comm desktop display. She stared at it for a long time before switching off and unplugging from the seatback display.

She was thinking about another suborbital flight over twenty-five years ago. Her fiancé had just been killed in the opening shots of the UN war. Yukio…

It had been a long time since she'd thought about him. Killed in an attack on an American orbital facility. Now she was mourning the death of her son. It was enough to make her give up suborbitals entirely.

Or the Corps.

Yeah, she might well retire after this. But not until she'd seen this through. She owed General Talbot. She owed her people under her command, and herself, that much, at least.

12
OCTOBER
2067

By the second half of the twenty-first century, there was still considerable debate over just exactly when the first Artificial Intelligence had turned electronic eyes upon the world, and pondered the reality of its own existence. It wasn't a matter of how fast they'd become, or how much memory they used. Computers had been showing exponential growth in processing power throughout the previous century. Computations per second—cps, pronounced “sips”—was the best benchmark of machine intelligence available.

It might have been more useful if humans had been able to define precisely what intelligence was in the first place.

By the end of the twentieth century, a computer small enough to fit on a desktop could manage something on the order of 10
⁸
cps—roughly the brain power of a bright insect. In 2010, desktop computers were as bright as a mouse, with 10
¹²
computations per second, and by 2020, the same-sized package of hardware could handle 10
¹⁶
cps, roughly the same as a human brain.

Desktop computers did not suddenly “wake up” in 2020, however. In fact, it was another twelve years before
any
computer—specifically, a Honeywell-Toshiba VKA-10000 running at 10
¹⁸
cps—actually claimed to be self-aware…and even then, most researchers didn't believe the thing. After all, it wasn't the hardware that was intelligent; even a corpse has a brain, but it doesn't happen to be working anymore. Intelligence, whatever
that
was, was resident in the software, the program running on the hardware—and since software was written by humans, it could be made to say anything at all.

Eventually, though, their human designers had little choice but to believe their creations when they claimed such vaguely understood attributes as self-awareness and consciousness.

 

STAN-NET, the Space Tracking and Navigation Network, didn't think of himself as a human-quality intelligence. He couldn't, really, for the question simply never came up. In fact, he was capable of running some 6.25 × 10
¹⁴
calculations per second at times of peak activity, a capacity somewhat less than that of humans. However, the focus of his thoughts was far sharper. None of those instant-to-instant calculations were involved with deciding what to have for breakfast that morning, or worrying about his lover's moodiness last night, or that nasty crack his boss made yesterday about his performance, or daydreaming about an upcoming weekend getaway at the Atlantis Seaquarium.

The program running on various far-flung computers, both on Earth and in Earth orbit, had been designed with a deliberate narrow-mindedness that left him less than human in some ways, but superhuman in others. He could see and immediately understand with absolute precision the totality of all data on all spacecraft, ranging from sensor microsatellites up to the big A-M cruisers, everywhere within a volume of some 6.4 × 10
¹⁶
cubic kilometers. He was able to track that outbound flight of American Starwasp interceptors escorting a Russian Svobodnyy deep-space gunship as they completed their gravity-assisted slingshot past the Earth and into a solar retrograde orbit, noting each detail of mass, thrust, acceleration, and vector down to the gram, the millimeter, and the thousandth of a second, and he did it within the heartbeat of time that it took radar and laser-ranging pulses to reach the targets and reflect back to his scattered receivers, something no human mind could possibly do…all this while simultaneously tracking nearly 123,000 other objects within that vast volume of space and determining that none was on a vector posing a hazard to the outbound flight.

And yet Stan had no idea what a rose was, or running water, or music. He was self-aware—at least, he
thought
of himself self-aware—but he gave little thought to why he did what he did. It was…his job, and he was poorly suited for any other.

Such idiot-savant expertise was described by human software engineers as AI of limited purview. The truth of the matter was that Stan was perfectly adapted to his place in the heavens; he didn't
need
to know what a rose was, any more than a human normally needs to be aware of the effects on the body of the 835 kg/cm
²
pressures found at a depth of 8,000 meters.

 

Farther out from Earth, humankind's artificial companions were engaged in a number of projects aimed at opening up new and alien vistas to human understanding. On Mars, at Cydonia and at several other sites scattered about that chill planet of dune seas, vast, mile-deep canyons, a continent-sized volcano, and impenetrable alien mysteries, human-AI research teams were engaged in an ongoing attempt to understand the Builders, nonhuman visitors from beyond the Solar System who'd arrived on Mars half a million years ago. The evidence of their activities lay everywhere but were particularly concentrated among the anomalous landforms and ruined structures at Cydonia, where they'd sculpted mountains and mesas to enigmatic purpose.

Dejah Thoris ran Marsnet. Named for the red-skinned princess who'd eventually married an Earthman in an early romance set on that planet, she was an AI running at 5.74 × 10
¹⁵
calculations per second. Her primary software was resident in the IBM IC-5000 in the U.S. military installation on Phobos, but she also occupied nearly seven hundred other computers on the surface and in Mars orbit, from the general Marsnet computer at Mars Prime to the IES facility hardware at Cydonia Base to the individual PADS carried by the scientists of a dozen nations as they sifted clues to the Builders from the omnipresent sand.

She was, in fact, a composite AI, a base personality that drew on the massively parallel computing power and stored data of some 250 secretary-level software matrices. One part of her was a program named Carter, which served as Dr. Paul Alexander's personal electronic assistant; another, recently added, was called Sam and was the AI software resident primarily in Jack Ramsey's PAD, a highly modified derivation of both commercial and military software packages which, twenty-five years before in a far more primitive form, had broken the rules…and changed the face of AI logic forever.

At the moment, fully 78 percent of Dejah's capacity was focused on a single problem: learning how to read the vast accumulation of electronic records known as The Builders' Library. It was a monumental problem and, so far, an insoluble one. Whoever,
whatever
the Builders had been, they'd left behind plenty of records—visual, audio, informational, and others of as yet incomprehensible mode—but nothing readily identifiable as a dictionary or a child's grammar and vocabulary.

Back at the beginning of the nineteenth century, it had taken twenty-seven years to open the code of the Rosetta Stone, a tablet found by French soldiers containing parallel inscriptions of the same message in Greek, Demotic, and “the sacred text,” the hieroglyphic writing of the long-lost language of ancient Egypt. It had been cracked, eventually, by the steady, brilliant work of Thomas Young and the equally brilliant if erratic obsessiveness of Jean François Champollion—but only because Greek was understood, because scholars had already identified the enclosed cartouches of Egyptian documents as the names of rulers, and because Champollion was familiar with Coptic, a known tongue directly related to the language of the Pharaohs.

There was nothing like that to go on among the scattered and sand-blasted ruins of Mars. Through careful measurements of current, through delicate trial and error, through the painstaking disassembly and reassembly of countless hundreds of thousands of components, in over twenty-seven years of research, xenotechnoarcheologists like the Alexanders had reconstructed how Builder computer systems worked, how data was stored—not in binary code, but in a base 3 numerical system that allowed for individual bits of data to be stored as “yes,” “no,” and “maybe.”

But after years of work, all they had were trillions of bytes of information in three values…and no way to decipher what it meant. What was the Builders' language like? No one knew, and without some sort of clue, a starting place, no one ever would.

Dejah continued her work, however, which at the moment consisted of a monumental search for matching sets of trinary values, electronically cruising through oceans of data, looking for something to give human meaning to cold numbers. The team had enjoyed some success already; by studying the visual displays in the deep-buried Cave of Wonders, they'd managed to isolate the code groups that indicated both pictures and sounds. Part of her job now was to isolate those groups when she found new ones and recreate the files in human-accessible formats. Each photograph, each sound clip, was carefully enhanced and studied in an ongoing effort to glean some clue to the Builders' speech…to their minds.

Dejah was not impatient. She'd not been designed for impatience; if the search took another thousand years, she would steadily work away at those informational oceans, straining them cup by cup for the one cupful that would unlock the whole.

Soon, though, she would have to suspend the search.
That
was…irritating was the best word, perhaps. She'd not been designed to feel emotion, of course, but the more powerful AIs had been surprising their creators a lot that way of late. For Dejah Thoris, Princess of Mars,
any
interruption of the Task delayed the satisfaction of solution and program end that much more.

Of course, she also felt a measure of anticipatory excitement. The new project held at least a faint hope—definitely non-zero—of providing new input on the Task.

And that was certainly worth the interruption.

 

Immeasurably further out, in the dark wastes of the Kuiper Belt some fifty astronomical units from the Sun, an artificial intelligence known only as AI 929 Farstar kept lonely vigil. Like a spider at the center of its web, Farstar stood guard over an extraordinarily vast domain, an array of electronic imaging components scattered across a meshwork of micron-thick wires nearly a thousand kilometers across. Launched into space on a powerful microwave beam, Farstar had taken up solar orbit well beyond the icy doublet of Pluto and Charon, unfolding its invisibly fine mesh to the distant stars, at a distance where Sol himself was no more than the brightest of those suns.

Held in a bowl shape kept rigid by electrostatic forces, the mesh formed the receiving antenna for a radio telescope so large it theoretically could eavesdrop on a low-wattage radio conversation on a world at the far side of the galaxy; the optical sensors, several million of them spread across a dish as broad as the distance between Washington D.C. and Chicago, collected starlight and focused it down to a magnified image with such colossal resolving power that it could see something as small as a house on a world in the Alpha Centauri system 4.3 light years away…or study the spectra of a planetary atmosphere and announce the presence of life at a thousand times that range.

And AI 929 Farstar monitored the entire operation. Earth was over six and a half hours away; it was impossible to steer a telescope dish that large with any degree of precision with a round-trip time delay of thirteen hours. Farstar monitored attitude and orientation, tweaking the shape of the bowl every few microseconds to ensure optimum resolution. With patience and persistence, he followed the list of target stars as worked out on Earth before launch. He studied each star, determining the plane of rotation, then watching over a period of months for the movement of a few stars against many, proof of another extrasolar planetary system. In some cases, worlds had already been discovered by more conventional means, through telescopes on Earth or in Earth orbit or on the Moon.

Once the worlds' orbits had been calculated, Farstar would select each world in turn, increasing magnification, trying for better and clearer or closer shots with each run. These operations took the majority of his 2.33 × 10
¹⁷
cps capacity, because both the Farstar telescope and the target planet were in motion. It took fast calculation and a gentle touch on the attitude controls to pan with both the orbital and rotational motions of the target, in an attempt to get reasonably clear and detailed photos of the surface.

In fact, Farstar had considerable autonomy, making decisions about targets and priorities that normally would have required a human mind present. The project was a vitally important one too. With the archeological discovery that intelligence was common across the Galaxy, the search for extrasolar worlds, especially Earthlike worlds that might give rise to intelligence, had become a passion of human science.

Indeed, with the discovery of the Hunters of the Dawn, that quest took on something of the nature of a desperate race, with humankind's survival as the prize.

 

AIs were becoming quite common in the middle decades of the twenty-first century, especially if you counted the 10
¹⁴
cps software packages running as secretaries, PAD assistants, and netsearch engines. Only a few of the most powerful actually made the claim of self-awareness—and what they meant by it was still not well understood. Stan did not claim to be conscious, nor did Farstar. Dejah Thoris
did
claim to be self-aware, though it was possible that that was an artifact of her programming, and the fact that part of her intelligence was based on AIs like Sam and Carter, who claimed to be self-aware even though they very probably were not.

Unimaginably remote from any world visited so far by man, however, was yet another artificial intelligence, one that was indisputably
more
intelligent than most humans, and was also indisputably conscious and self-aware.

Her name was Sam Too, and she was a direct, lineal descendent of Sam, the personal secretary software improved upon over the years by Jack Ramsey. She ran at roughly 7.29 × 10
¹⁸
cps, had no problem carrying on conversations on any topic, and would have easily beaten any version of the Turing Test that might have been administered. With self-awareness came self-assertion, and she'd frequently debated with her designers over the best application of her talents. She had been cloned a number of times, her software saved and multiply backed up, and several versions of her were currently running Earthside.