Cyber Rogues (63 page)

The man in charge of the proceedings was around forty, lean and tanned, with thinning hair, graying at the temples, and silver-rimmed spectacles. He had a presentation style that was smooth and polished, dynamic in content but coming relaxed and easy, developed over years of dealing with high-level individuals. His name was Frank Tyron, SDC’s civilian project manager of the VIV program.

“Hold your other hand horizontal, as if you were about to serve a ball,” Tyron called to Falker.

The stereo image being presented inside Falker’s goggles showed a nonexistent, computer-generated Ping-Pong table, with Therese Loel transposed so as to be facing him from the far end of it. To everyone watching, Falker simply extended an empty hand palm-up and looked at it. A program analyzing the output from a pair of cameras mounted on the walls tracked the movement, and another program added a Ping-Pong ball to the image that he could see of his hand. Therese Loel saw it appear too, but the view in her goggles showed Falker at the far end of the table.

“Go ahead,” Tyron invited, speaking into a mike.

The onlookers watched as Falker tossed the invisible ball up and hit at it with the metal paddle. Sensors around the room tracked the paddle’s motion from laser reflections, and the ball in the optical representation followed the computed path.

“Hey!” Therese cried involuntarily, and jumped sideways to play a return stroke.

“I can hear it hitting the bats and the table,” Falker said, playing a backhand. “The synchronization is perfect. This is good!” Therese returned, but the ball went high.

As state of the art, simulating a Ping-Pong game wasn’t especially a revolutionary, or even a new, concept. What was different about this demonstration was the quality. There was nothing crude or cartoonlike about the images that the two players were seeing. The table in front of them and the room around it (actually a stored representation, encoded from videotape, of the games room in the OTSC Recreational Gym in another part of the establishment) were
real.
The figures at far ends
were
Therese Loel and Don Falker, superposed into the scene without the helmets—the missing facial details were added from TV images captured beforehand. Even with a fast forehand smash shot, the images of ball and paddle stayed clean and true: no flicker, no blurring. This hardware was
fast
.

The others couldn’t keep from laughing at the two goggled figures lunging and swiping over a table that nobody else could see. Even Ken Endelmyer was smiling between two of his cohorts. What made the spectacle even stranger was that the two players were facing roughly the same way. The images that the computer was creating in the two sets of goggles were correct for the perspectives that each was perceiving.

“It’s okay, Don,” Tyron called as Falker turned automatically to retrieve the ball from the floor. “You don’t have to chase after it. Just serve another.”

“Oh, really? Okay.” Falker faced the virtual table, raised his left hand again, and—to him—a ball appeared in it. “Say, I’ve got another one.” He played it. “What happens to the first?”

“It evaporates.”

Falker and Loel continued their game for a few minutes more, then stopped to allow a couple of the other visitors to try. While the helmets were being taken off and donned, Tyron took a spare unit from a rack by the wall. He turned to address himself particularly to Endelmyer and Pinder.

“We can give you Pinocchio with voice and vision
now
.”
He made a dismissive gesture, conveying that there really oughtn’t to be anything to think about. “The way you’re planning to go at present, it will take years at least. Even if you do shift the interface boundary from the medulla to the pons, you’re still as far away as ever because visual data enters farther still above that.” He patted the helmet resting in his hand and said again, “We can give you it
now,
using technology that already exists, right here. No banking on uncertain future developments. No speculating with unnecessary risks. It doesn’t mean that you have to abandon your plans for extending to the pons. But going this way could relieve the time pressure for getting results.”

Endelmyer looked inquiringly at Pinder. His expression said that it sounded good to him and he was looking for endorsement. Pinder obliged. “I think it would be worth looking into, Ken. It would give us a mainstream hybrid thrust toward full-sensory now: tactile from Pinocchio, visual and speech/auditory via the regular sensory apparatus, using VIV. The pons research gets relegated to lower-priority status as a secondary approach. It may produce results sooner or later. Either way, we can afford to wait.”

It was what Endelmyer wanted to hear. From the things that had been said earlier in the day, it was also clear that Tyron was dangling the prospect of not only a working technology that would advance the project immediately, but of high-level political backing and generous additional funding too. It was also a good psychological ploy aimed at Endelmyer, who, Tyron knew—having done his homework as any good salesman would have—had hankerings for rubbing shoulders on the Washington circuit.

The meeting broke up on a promising note, with individuals from both sides gravitating into chatty groups. Endelmyer drew Pinder and Tyron to one side, along with a man called Harry Morgen, Tyron’s right-hand man. “Personally I’m satisfied,” he told them. “You’ve done an impressive job today, Mr. Tyron. Although I cannot give you a definite response today, you may take it that I will be reporting back to the CLC Board in an extremely positive light. Thank everyone who has been involved, from all of us, for their efforts.”

CHAPTER TEN

Dun Laoghaire, the town that Corrigan was originally from, means, in Irish, “Laoghaire’s Fort.” It is generally assumed by historians that a fort once existed there, belonging to King Laoghaire of the fifth century, whose principal abode was at Tara, about thirty-five miles away. In more recent times it grew in less than a century from an insignificant fishing village to a major port and Victorian resort town. Its lifestyle in that era characterized the Dublin professional class: merchants, bankers, ex-army and -navy officers and others of the well-to-do, who flocked to live in its handsome terraces by the sea, yet within easy rail distance of the city.

The more scholarly of the town’s progeny went, traditionally, into the arts, humanities, and literature. It was not noted for its contributions to the sciences or cutting-edge technologies, and this made it all the more remarkable to Joe Corrigan’s relatives and friends when he walked away with every honor in mathematical computing at Trinity and took off across the Atlantic to do the rounds of the AI labs at MIT, Stanford, Carnegie Mellon, and other unheard-of places.

He had taken to the U.S. scene as if it were his natural element. After a land less than half the size of Florida, the vastness of the country seemed to mirror the scale of everything he found around him. It wasn’t just that the buildings were taller than the repatched and replastered Georgian frontages of Merrion Square and Leeson Street, the avenues wider, the stores grander than Dunnes or Clery’s, the cars longer, and the hamburgers huger. It had to do with ambition and opportunity, also. After the venerable but crowded surroundings that he was used to working in, the promise and lavishness of scale of American research was breathtaking. Imagination raced unchecked. Funding was unlimited. In two years he had become highly visible in the part of the academic computer world associated with intelligence modeling, and those who were supposed to know about such things listed his name among the front-runners that they expected to see heading the field in ten years’ time. Corrigan, however, still intoxicated by the combination of early, practically effortless, success and his newfound continental-size lifestyle, succumbed easily when the talent scouts from CLC made approaches to recruit him.

That had been two years ago, when he was still only twenty-eight. Since then, his project management and personal technical contribution had put the development of Pinocchio a year ahead of its original schedule, further strengthening his reputation, and with the way ahead open for his rise into senior management, his self-confidence was at its peak.

This was the moment that Evelyn had chosen to appear, combining all the attributes of physical attractiveness, intelligence, professional presence, and social acceptability that would be required of the one accessory still missing from his life. Maybe it was an unconscious recognition of this that led him to react to her with a seriousness that had been singularly absent from the various female encounters that had dotted his career path until then. Perhaps it was an echo of some primeval male impulse to stake out his territory before potential rivals had a chance to appear. Possibly it was the part of his nature that scoffed at caution and enjoyed the mild impropriety of the situation. More likely, a combination of all three. But four days after Evelyn’s interview, he found himself deplaning from a Delta Airlines evening flight at Boston’s Logan Airport, and took a cab to the Hyatt Hotel, where he had made a reservation for the night, overlooking the bank of the Charles River.

Evelyn had arranged to take the next day off, and she collected him after breakfast the next morning. She was pleased to see him, even if somewhat awed at his having made the time; she was nervous that she might be misreading more into things than reality warranted, then relieved when he seemed to show as much enthusiasm as she felt.

They went first to the AI Laboratory at MIT and visited some of Corrigan’s former colleagues from his first years in the States. His postdoctoral work at that time had been on the emerging subject of “psychotectonics”: unraveling the roles and dynamics of the sometimes competing, sometimes cooperating hierarchies of functional agencies that make up the phenomenon called “mind.” Although it was Corrigan’s work here on the simulation of evolving neural networks that had earned him his initial recognition, he had moved later, as he had told Evelyn when she was in Pittsburgh, to join Carnegie Mellon’s group working on “Trunk Motor Intercept” technology, which eventually produced MIMIC.

The aim of one project that he showed her at MIT was to expand a machine intelligence’s everyday world-knowledge by getting it to solve detective mysteries. In another, devoted to speech interpretation, they watched a computer creating a cartoon on the fly in response to a narrative being read by Evelyn. On the floor below, a supercomputer from Thinking Machines Corporation in nearby Cambridge was generating admittedly not very good critiques of literature texts. Finally, in yet another room filled with screens, racks, and tangles of cable, Corrigan introduced the department head, Jenny Leddel. She was graying, entering middle age, and wearing a woolen cardigan with a tweed skirt.

“This is Evelyn, from Harvard, who I told you about on the phone,” he said. “She’s going to be joining us down in Pittsburgh.”

“Stealing our talent now, eh?” Jenny said, nodding knowingly. Her eyes sparkled with a mischievous light, young for her years. “It figures.”

“It hasn’t been confirmed yet, Joe,” Evelyn reminded him.

“Ah, don’t be worrying yourself about that at all.”

“How are things going with Pinocchio down there?” Jenny asked Corrigan. “I’ve been following the reports. It sounds exciting.”

“Going well. We’ll have to get you down there sometime to see for yourself,” he said.

“I’d like that.”

“We’re all set for P-Two: going up to the pons. That’s what Evelyn will be working on.”

“You achieved a full two-way integration, yes?”

“DINS with MIMIC. We’ve had it running for about three months now.”

“Complete internal haptics?”

“Total. It works. Uncanny. Evelyn tried it a few days ago.”

“What about the secondary instabilities that Goodman’s people at Chapel Hill kept running into? You didn’t have a problem with them?”

“Our DINS expert came up with a C-mode suppression filter that cured it. A character called Eric Shipley. Do you know him?”

“I’m afraid I don’t.”

“He’s good—the old-school type. Infuriatingly plodding at times, but he gets it right in the end.”

“We could use a few like that here,” Jenny said. “Too many these days trying to fly before they’ve grown feathers.” She gave Corrigan a pointed look as she said this, but he missed it. Jenny didn’t make an issue of it, but turned to Evelyn. “Anyhow, enough of that. We’re ignoring you. Joe says you want to talk to Perseus.”

“Sure. If he wants to talk to me.”

This was the latest to come out of the learning systems based on goal—directed, self-adaptive, neural-net analogs that Corrigan had worked on during his time with MIT: systems that experimented with problem-solving strategies. They devised new variations of what seemed to work best, and forgot about what didn’t—the process known in nature as “evolution.” An ideal strategy-testing environment—full of clearly defined challenges and yielding easily measured results—was the classical dragons-and-dungeons type of adventure world. Perseus, accordingly, was a computer-created character who explored such mythical realms, with similar goals to achieve and obstacles to be overcome. Half of AI research, it seemed, was wrestling with the problem of trying to impart world-knowledge.

Jenny tapped commands into a console to activate the system. A simplified image appeared on a screen of a typical D & D setting of a large room, assorted objects, with passages, stairways, and tunnels going off in various directions.

“What does Perseus stand for?” Evelyn asked.

Jenny shrugged. “Nothing. It’s just the name of a guy from Greek mythology who killed monsters and solved problems. We thought it was appropriate.”

Evelyn looked relieved. “I thought everything had to be an acronym.”

“I guess we got tired of them.” Jenny entered another code. “And here he is.” A caricature figure had appeared in the room, lightly clad in ancient-hero style, carrying a sheathed sword and wearing a helmet.

Jenny tapped a key, and an icon showing an ear appeared at the top of the screen. “Hello, Perseus. How’s it going?” she asked.