Coming of Age: Volume 2: Endless Conflict (16 page)

“I’m going to note the defects which the clerk identified in the prenuptial agreement, but I’m ruling that they don’t invalidate it. So the agreement stands.” She paused. “I take it that the defendant does not wish to challenge the divorce itself?”

“No, your honor,” Wells said. “Only its terms.”

“This is what you want, Ms. Jaspersen?”

“I do, your honor,” Rafaella said.

“Very well. The decree stands. As to the terms, I will confirm Ms. Jaspersen in custody of the three daughters, with Mr. Jaspersen to have visiting rights twice a month and for two weeks per year, dates and times to be worked out between the parties. He will also pay alimony and child support.” The judge consulted Rafaella’s petition and named a sum that would let the girls continue in private school but—by Rafaella’s quick calculation—require them to move to a more modest apartment. And that was fine by her.

“Anything else?” Judge Biggs asked.

“Your honor,” Tim’s attorney began, “I must protest!”

“Yes, Mr. Averill, I thought you would. … I don’t know how this bit of judicial malfeasance came about, especially your client’s representation of how the terms of his employment came to be jeopardized. I don’t like to impute perjury when a simple misunderstanding may be at fault. If you want to pursue the matter, you can move to vacate this judgment—but I’d advise you against it.”

With that, she banged her gavel, bringing the case to its close.

Rafaella was free, and her—and her girls’—future was assured.

* * *

“Hello, Cousin,” Jacquie Wildmon said when Brandon Praxis’s image came up in the left-hand field of her office’s communications wall. He was looking pretty good for a man who was now—what? Almost sixty-three years old. His face might have been that of a man in his forties. The right-hand field came alive with the image of a young woman with curly reddish-brown hair, alert blue eyes, and snub nose. Brandon’s wife and business partner, Penny. “Or ‘cousins,’ I should say. What can I do for you?”

“We need your help,” Brandon began, waving vaguely to include Penny, who was obviously in another place and time zone. “We’ve had an issue with one of our Watch and Ward
^®
intelligences.”

“It appears to have made a mistake,” Penny continued for him, “and then the software—well, lied about it.” Penny paused and looked as if she didn’t herself believe what she had just said. Brandon, on the other screen, simply nodded and shrugged.

“For gosh sakes!” Jacquie muttered. “I’ve become Susan Calvin!”

“Who is that?” Brandon asked, and Penny stared across at him.

“The robot shrink in the Asimov stories,” Jacquie explained.

Her position as Tallyman’s expert on “AI Developmental Problems,” meant she had become a cross between computer programmer and child psychologist. One of the issues she had to deal with was the question of intention and ethics in minds that ranged so far and wide, that processed information so quickly, and that had to deal with sometimes imperfect data. Could an AI develop or discover for itself an ulterior motive? Could it choose to do the wrong thing or make a mistake and then give false witness? Could it become bored and mischievous?

Sometimes Jacquie wished that every problem could be solved with a syllogism as simple as the “Three Laws of Robotics.” Her kids used to ask her about that as soon as they started to read science fiction. First, a robot must never harm a human being or let one come to harm. Second, it had to obey all human orders, unless they conflicted with the First Law. And finally, it had to protect itself, unless that action conflicted with the First and Second Laws. Something vaguely like this syllogism was built into the core programming of all intelligences—but more as a guideline than as immutable law.

“Those laws will work—most of the time,” she had told her children, when she thought they were old enough to understand. “They apply to first-order actions, anyway, like crashing an airplane full of human beings or wiring an electric toothbrush at a thousand volts. And we do give AIs a grounding in safety parameters, both for people and for themselves.”

But life, for a machine as well as for humans, was full of complexities, shades of gray, and mazes of interlocking decisions and choices. The end-result of most choices could no more be foreseen and resolved than the average person was be able track the results of a chess problem twenty moves into the future—which most AIs could do with ease. And, too often, too many optional strategies and pure-choice moves depended on subsequent actions of the opposing player. The machine choice that might avoid “injuring a human being or, through inaction, allowing a human being to come to harm” was pretty clear when A shoved B in front of a train—but less so when weighing choices in design of the train’s air brakes. In every situation and transaction—whether in engineering, economics, medicine, transportation, or the thousand and one other realms of activity that machine intelligences now controlled—important and necessary tradeoffs existed. To write a simple rule, or try to elevate a single consideration like “human safety” or “customer satisfaction” or “system protection” as an operational absolute was to invite paralysis and gridlock. The world was just too damned complicated.

“So tell me your story from the beginning,” she told her cousins.

And they proceeded to describe a Watch and Ward
^®
security system that had appeared to confuse registered visitors who had approved access to a restricted area during construction with a whole other class of operators who were not due inside that same area for another two years. When confronted with the problem, the W&W brain had claimed corruption of its database—something like “garbage in, garbage out.” And then, when confronted with its own clean and tidy, uncorrupted database, it had claimed some form of unexplained amnesia.

“Do you still have the intelligence online and at work?” Jacquie asked.

“We …” Brandon stopped and turned toward his physically absent wife.

“It’s … still on the job,” Penny acknowledged. “It didn’t actually harm—”

“I see. And that’s good, because then I can interview the brain in real time.”

“I’ve already done that,” Penny said. “A data dump and review of its structural code. Everything checks out, as far as I can see—except for that lie about not remembering.”

Jacquie thought for a moment. “Can you think of any reason why someone would want to manipulate its access roster? Maybe to allow bogus visitors on site? Steal tools or plans? Commit sabotage?” She was remembering the other problem the Praxis family had recently brought to her, about Rafaella’s divorce and its lopsided judgment, where an ulterior motive was certainly possible if not probable.

“Not with these visitors,” Brandon said. “One was a client, the other a local inspector. Getting them onto the site was absolutely necessary to the day’s work.”

“Well, I’ll still want to talk to your—what was the name? ‘Officer Krupke’?” Jacquie laughed at that. “I’ll grill him a bit, give him the old good cop, bad cop routine.”

Penny looked dubious. “I don’t know what more you can do than I already—”

“Right,” Jacquie interrupted. “But understand that I have better tools here.”

“Won’t the machine just keep lying about what it’s done?” Penny asked.

“You mean, create a false record, a fictitious dump, more error?”

“Yeah, like that. In my experience, these things are
smart.
”

“And some are smarter than others,” Jacquie replied. “If you want to catch a child out in a lie or a petty theft, you ask it directly. And when it denies everything, what do you do then? Do you ask another child? No, you send in an older, more experienced person to question him and examine his responses.

“Here at Tallyman,” Jacquie went on, “we have some of the smartest intelligences on the planet.” She was thinking of using Vernier, her partner in machine analytics, for this case. “They can analyze the pattern of original errors, compare it with the AI’s internal dump and subsequent responses, and tell us exactly what happened. My machines cut their teeth on unequal data sets.”

The cousins looked dubious, but in the end they gave her Krupke’s online address and backdoor access code. This was going to be fun.

* * *

Before he would give Susannah a firm answer about her ideas, John Praxis wanted to discuss them more generally with some of the senior family members. He called them into his office at the engineering company headquarters—Callie, as chief executive of their construction business; Jeffrey, who managed the new forest lands subsidiary; and Brandon and Penny, who had branched out into automated security systems.

He explained his great-granddaughter’s idea for what he chose to call the “Praxis Family Association,” which would be something like a cross between a clan, a family business, and a commune. The details were still hazy in his own mind, but he grasped what Susannah had been driving at: a path for survival and sustainability in the generations to come. Where these various family businesses would lead them, what they would try and either discard or succeed at, and how they would survive—all those details would be shaped by events and opportunities. The only limiting factors would be the quality of their own wits, the power of their collective vision, and whatever obstacles and restrictions the State of California in Sacramento, the Federated Republic in Kansas City, or the International Court of Justice in The Hague might impose. Everything else was up for grabs.

Callie’s frown grew deeper as he talked. When he finished, she asked, “How are we going to manage this?”

“I assume it’s going to be structured as some form of corporation with subsidiaries,” Praxis said. “Much the same as now—except that every family member will have a share.”

“Equal shares?” his daughter asked. “Some of us have more at stake than others.”

“Well …” He considered. “We should have some form of proportionality—at least to start. Recognizing the varied contributions of each member.”

“Would we still have private property?” Brandon asked. “For instance, would Penny and I own anything for ourselves, like what we’ve built up in Watch and Ward? Or would it all be given over to this new super-company?”

“And would those shares be inheritable?” Callie asked. “I have one daughter. Brandon and Penny have two children, a son and daughter. That would tend to concentrate my holdings—”

“Only until the next generation!” Penny protested, then stopped. “Pardon me! I don’t know if I’m allowed to speak—as an outsider.”

“Of course you can, dear!” Brandon said quickly.

“I’m only an in-law,” his wife replied.

“Go ahead,” Praxis told her.

“Well, Rafaella has three daughters,” Penny said, “so in that generation whatever shares you bequeathed to Rafaella would be diluted again. It seems to me that having more children would then become a disadvantage—like the old landed Irish, cutting up the family farm into smaller and smaller plots for the sons to inherit—rather than the strength that sons and daughters ought to be.”

“What about a birthright?” Jeffrey asked. “There would be inherited shares, sure, but each new child—and each person brought into the association by marriage or adoption—they ought to get something right away.”

“I’m still concerned about that original distribution,” Callie said. Praxis could see that, as his only living descendant in the first generation—other than Alexander, but his son with Antigone was something of an anomaly—Callie would feel she had something to lose. “Will it be by seniority?” she went on. “Amount of contribution? Degree of kinship? For example, does a third cousin by marriage get as many shares and as much influence over the operation as, say, I do?”

“That raises an interesting question,” Jeffrey replied. “I’m back inside the family business now, working for you—but what about someone like Jacquie, down at Tallyman? She has a career outside the family business. More to the point, sometimes Tallyman is our supplier, but more often our competitor. So … does she get roped into this thing or not?”

“She’s Richard’s daughter,” John said. “She’s a Praxis. So, yes, she’s in.”

“But how do we account for her share without raising a conflict of interest?” Jeffrey pressed. “And if she refuses to join, do we have to compensate her?”

“Excuse me, John,” Penny said hesitantly. “I think we want participation and ownership to be as wide as possible. That’s the way to build up our base—both for personal contributions to our productive capacity and for potential market demand for our products. This thing won’t work if it’s held too tightly.”

“I agree,” Praxis said. “We want to run a big tent.”

“More mouths to feed,” Brandon observed.

“Not a problem,” Praxis said. “What with automation, a good base of capital, and sufficient raw materials, we’ll be rolling in products—or at least that’s Susannah’s analysis. The difficulty will not be production but consumption.”

“Come again?” Brandon said.

“Something she said struck me,” he explained. “Since the Industrial Revolution, and even more so since the Robotics Revolution, we’ve turned the economics of human experience on its head. Consider a hundred thousand years of hunter-gatherer history, or the first couple of thousand years of settled agriculture. What defined the good times? Big harvests, full granaries, lots to eat. Bad times were drought and famine. But with the productivity that science and technology have brought to agriculture, and to every other sphere of industry since the Enlightenment, what are the good times? Lots of demand pulling through the system, assuring full capacity and growth. Bad times today are defined by failure of demand, inventories piling up, plant closures, layoffs, recession, depression. Whether we invest in a fully automated factory or just a three-D printer, the issue will be keeping it busy and pulling that demand.”

“So … big tent, lots of shares,” Callie said. Still, she looked dubious. “But what about contributions? Do we have a work requirement? Or does everyone get to eat for free, while you and I do all the work?”

“The bigger question,” Praxis said, “is how to keep our people fulfilled, happy, and hopeful. For the last six months, Callie, when Rafaella’s future was up in the air and you had to support her, do you think she was happy?”