When HARLIE Was One (10 page)

Carl Elzer looks like a weasel
.

Auberson knew the thought was unkind—but that still didn't keep him from thinking it.

I have a right to be cranky
.

The boardroom smelled of old leather, stale smoke, and cologne so sweet it made Auberson think of rotting flesh. He did not like meetings with the board; he resented the time spent away from the real job. HARLIE.

But the board meetings were part of the job too. They were the occupational hazard.

I
have no friends in this room.
Auberson realized he was fingering the pencil in front of him. A few more minutes of Elzer's prattle and he'd start doodling. He pushed the pencil back in place next to the pad of paper and pushed both away from him. Out of temptation's reach.

The board was an unequal mix of stiff old men and narrow-eyed sharks. The stiff old men had skin like parchment. They sat impassively and watched, and you couldn't tell what they were thinking. The sharks just circled patiently, waiting for the scent of blood in the water.

Carl Elzer was neither a shark nor a stiff old man. He was a weasel. A ferret. A verminous little rodent with the morals of a piranha.

Goodness, I am getting cranky
.

Elzer was reading from a voluminous sheaf of notes in front of him; something about the balance of cash flow to research, and how the company's ability to invest in research that would produce immediate results was being hampered by persistent long-term drains on the operating capital. He rustled his papers importantly, then looked across the big conference table at Auberson.

Here it comes,
thought Auberson.

“Now, then—that brings me exactly to the question at hand. As I understand it, you've had something of a setback. Isn't that right?”

“Actually,” said Auberson, very quietly, “What we've had is more in the nature of a breakthrough.”

“I beg your pardon?” Elzer blinked.

“I said, we've had a breakthrough.”

Elzer made a show of sorting through his papers. “I'm afraid that I don't see any, uh—evidence of that here.” He sorted a moment more, then lifted his gaze again. “Is that something that happened just this morning? Or—what?”

“I think we're talking about the same set of events,” said Auberson. “It's the interpretation here . . .”

“Ah! I see.” Elzer put his papers aside politely. “Yes, I'd be very interested in hearing your explanation why this is a—what did you call it?—a ‘breakthrough'?”

Like hell you would. You just want to hand me enough rope for a hanging
—

Auberson leaned lack in his chair and studied Elzer. It was a carefully practiced maneuver—almost a reflex. He used it whenever he was uncertain of how to proceed. He decided to put the ball back in Elzer's court.

“Why do you think it
isn't
?” he asked.

Elzer looked up and down the table with an annoyed expression on his face. “You know, this just proves what I've been saying for years. Nobody really reads these reports.” He brought his gaze back to Auberson, adjusted his glasses on his nose, and said, “It seems to me that you are farther from producing a result than you ever were before. This whole thing
is
about results you know; but your machine has been, ah—there's no polite way to say this, Auberson, so I hope you'll understand that there's no offense intended—but in the past six weeks, your machine has been going into failure mode on an almost regular schedule; and I understand that the rate of these occurrences has been growing rather than decreasing. This does not sound very much like a breakthrough to me.”

“These ‘occurrences,' as you call them, are periods of nonrationality.” Auberson corrected. “But it's incorrect to call them a failure mode—”

“You called them that yourself—” Elzer said, shuffling through the papers in front of him. The assistant weasel on his right slid a manila folder sideways to Elzer. “Ah, thank you, Platt.” Elzer focused through his bifocals. “It says so right here, in your report of—”

“I know what I wrote. I was
wrong
.”

Elzer sat back in his chair. “You were wrong?” He blinked in surprise.

“Yes, I made a mistake. Does that surprise you? I know it doesn't happen very often—”

Elzer waved a hand. “On the contrary, I'm surprised to find you admitting it so honestly. It's a refreshing surprise. What else might you be wrong about?”

“We're getting off the subject,” Auberson said stiffly. “We were talking about HARLIE.”

“We were talking about
its
failure,” corrected Elzer. “You were about to explain why this was a . . . What was that word again? Oh yes—a ‘breakthrough.' Let me ask you something, Auberson. Is this the kind of breakthrough that we're going to be able to take to the bank? Or is this one of those ‘personal transformation' breakthroughs? We can't declare much of a dividend on those, you know.” Elzer put his hands together in front of him and steepled his fingers. He smiled solicitously. “Let me guess. You
recontextualized
the
process
—right?”

Auberson looked annoyed. He looked up the table to the president of the company. Brandon Dorne was a heavy-set man who sat quietly in his huge leather chair with his hands folded across his paunch, watching the verbal gunplay across the boardroom table. There was no help there. Auberson turned back to Elzer. This was going to require something drastic—

He cleared his throat. “Your ignorance is showing again, Elzer.” He said it softly and without emotion.

Elzer gaped. Before he could say something else, Auberson continued quickly, “What we're talking about goes a little bit beyond a bookkeeper's ability to conceptualize. What we're talking about here actually requires some real
thinking
.”

Elzer shut his mouth quickly. He opened it again, then shut it again. He looked like a fish sucking scum off the aquarium's glass.

Auberson continued calmly. “When I said breakthrough, I meant breakthrough. I'm talking about a development so astonishing that most of us in the department still haven't had time to assimilate it ourselves. Uh . . . it's not something that's easily explainable, but it has a very real effect on the direction of this project.” Auberson looked around the room. Was there real interest on the faces of the directors? Or was he just imagining it because that was what he wanted to see?

“The possibilities here,” Auberson said carefully, “are enormous. Much more than we considered when we chose our original goal. What we achieved, however, is the first glimmering of something so much more
powerful
that we would be stupid not to press ahead with this research as far as we can. We should not be talking about cutting back the HARLIE project, we should be talking about increasing his—”

“His . . . ?”
interrupted Elzer.

“Yes. His.”

“I see. Tell me, are you anthropomorphizing a personality into this machine, Auberson, or—are you trying to tell us that it has finally come to life?”

Auberson hesitated. How best to phrase this answer?
The hell with it, tell the truth
.

“As a matter of fact, that's exactly what has happened. HARLIE
has
come to life.”

Elzer stared. His eyes were narrow and hateful. “This is not a very funny joke, Auberson.”

Auberson stared right back. “I'm not joking. HARLIE's periods of nonrationality only looked like failures because we didn't know how to interpret them—”

“Stop it, Auberson. Just stop right there. The thing is silicon and electricity, gallium arsenide and lasers, clock-crystals and diodes and magnetism. You've said over and over that we can track every cycle of its internal workings. Now, you're trying to tell us that what was previously a comprehensible process has suddenly transcended that comprehensibility and become
life
?”

“Yes, that's exactly it. The fact of the matter is that we have succeeded—far more than we ever expected to. We only expected to simulate life. Unfortunately, we've simulated aliveness so effectively that we have no way of telling if it's a simulation any more or the real thing. And it doesn't matter. If Alan Turing were here, he'd tell you the same thing—HARLIE is alive because we can't prove that he isn't!”

“Who the hell is Alan Turing? And why
isn't
he here? Is he on our payroll too?”

Auberson suppressed an urge to giggle. Instead, he said simply, “Alan Turing was a World War II computer scientist who postulated many of the foundations on which the whole field of Artificial Intelligence is based.”

Dorne, sitting at the head of the table, took the cigar out of his mouth and said, “Let me get this straight, Aubie. You're saying that we've achieved a true Artificial Intelligence?”

Auberson nodded. “I think so, yes.”

“You
think
so . . . ?”

“We're still testing.”

Elzer remained unimpressed. “It's still a computer, isn't it?”

Auberson looked at him, frustrated by the man's inability to understand. “So is your brain, Elzer. Your brain is a computer made out of meat. In your case, hamburger. But it doesn't matter whether the software is running in meat or silicon. It still runs. The biggest difference is that silicon tends to be more accurate.”

“That kind of rudeness is uncalled for, Auberson.” Elzer snapped.

“Aubie—” Dorne said warningly.

“I'm sorry, I did not mean the comment as an insult. I was trying to make a point. It's a shame that Carl Elzer had to make such a pointed comment about the jargon of consciousness technology, because we've gotten some of our most interesting insights from some of the
est
holes and lifespringers and modies we've talked to. For instance, the
Mode
people call a human being a ‘self-programming problem-solving device.' That's also a good description of HARLIE. That's the essential goal of any artificial intelligence program—to have a machine that can understand questions in English and build its own programs for finding the solutions. Human beings do that too—except most of us don't understand the process by which we do it. HARLIE does—or at least he's trying to. That's what those periods of nonrationality are all about. He's trying to . . .” Auberson trailed off, abruptly unwilling to complete the sentence. He didn't want to hand new ammunition to Elzer.

“To what?” Elzer demanded.

Auberson sighed. “I think . . . that he's trying to understand what it is to be human.”

“Now you're saying he's not only alive, but
human
?”

Auberson shook his head. “Not human. Not exactly. How many human beings do you know who are immobile, who never sleep, who have twenty-five sensory inputs, who have eidetic memories, who have no concept of taste or smell or any other organic chemical reactions? How many human beings do you know who have no sense of touch?
And no sexual outlet?
Please, don't make that mistake. HARLIE is definitely
not
human. Not at all. But he does want to understand humanity—because we're all he has to talk to.” Auberson shrugged. “He does imitate human behavior, quite a bit. We're his role models. He's not perfect—or maybe I should say he
is
perfect. His emulations of us are terrifyingly accurate. So he makes perfect mistakes.”

“I beg your pardon?”

“He's developed a personality—quite a volatile one, I should say.”

“Volatile?” The little man was confused. “You mean he gets angry?”

“Angry? No, not angry. He can get impatient though—especially with human beings.”

“This is all very . . . interesting,” Elzer dismissed it with a wave of his hand. “And I suppose it's even quite a bit of fun to have all of these wonderful conversations with the machine. But it's all irrelevant. It's worthless research. Because it doesn't produce anything. The real question is—
when
are we going to see a tangible result from this project, Auberson? What is its purpose?”

“It's funny you should ask that. HARLIE has asked the same question. He wants to know why he exists, why
we
exist. Funny, isn't it?”

“No, I don't think so.”

Auberson ignored that. “The best I could tell him is that it's evolution. Intelligence is the logical result of the first law of biology.” To the questioning expressions around the table, Auberson explained. “The first law of biology is survival—”

“It's also the first law of business,” retorted Elzer. “Perhaps you should explain that to your machine. Or is it more involved with the existential side of the question? Perhaps we should have Miss Stimson, the executive secretary, arrange to bring in a minister to sit and speak with the machine.” A few of the board members smiled, but not Miss Stimson. She was quietly typing the minutes of the meeting into a portable terminal. “What we want to know is HARLIE's purpose. Having built him, you should have some idea.”

“HARLIE's purpose—I thought I just told you—is to be a self-programming, problem-solving device.”

“I mean for what
financial
reason? What economic applications will this program have?”

“Huh? The applications are endless. This is one of those developments that has so much potential—”

“Spare me the speech. Name six.”

“I can't just—Well, robotics, for instance. We could be talking about robot diagnosticians or negotiators, truly intelligent civil servants—or, controlling the right mechanical body, robot firemen or even police officers.”

“Robot police?”

“Or even soldiers or anti-terrorism squads. Any place where the ability to make quick judgments is needed.”

“Now, that might be worth something. How long till you could demonstrate a working prototype?”