[ Pobierz całość w formacie PDF ]
could not be sure. Not yet. Not until she got a look at the fiftieth SPR.
The fiftieth.
What bothered Fredda was the fact that she had not even noticed that an
SPR was missing. There had been fifty SPRs to start with, but she had not even
thought to do a count on them, until now. Now she knew there had been twenty-
two SPRs on the upper level, and twenty-seven on the ground floor.
If that information had been in her head earlier, she would have turned the
place upside-down to find that missing fiftieth robot. She would have found
this one, the crucial one, much sooner.
Not that this one had been overlooked by anyone except Fredda. Gallingly
enough, search teams had even logged in the location of this one two hours
ago, but they had not examined it closely. What was one more shot-up robot in
a building that was full of them?
She wanted to dive right in, to take this robot apart and find the clues, the
proofs, she knew were inside it. But she resisted. Suppose she set to work now
and smudged a fingerprint or something? No, thank you. There was no point in
making any more mistakes.
It had been frustrating enough to have that imaginary door in the integrator
simulation close on her face. To track the suspect this far and then come up
with nothing--that would be slamming into a wall. It was starting to dawn on
her just how much patience police work involved.
So, do it right, do it carefully. The clues in this room might be the core of
the case. Don t ruin them. Let the robots do their job first. Then she could
do hers.
Donald, Fredda said. I want you to call in a full team of Crime Scene
robots. I want this robot and this entire room--and all the storage closets--
scanned down to the maximum resolution. Our friend Mr. Pale-man was hiding in
here, and he must have left some traces.
That is by no means certain, Donald said. It would be most useful, but we
cannot count on it.
But he must have left something behind, Fredda protested. A bit of hair, a
fingerprint, something. Or was it possible that he could have left no trace
at all behind? Fredda suddenly realized just how little she knew about the
sort of clues she was counting on the robots to find.
It is possible the Crime Scene team will find something, Donald said, but
bear in mind that if our suspect took a few simple precautions there would be
nothing for us to find.
Precautions? Fredda was suddenly confident of her ground. Forensics and clues
she did not know about, but people she understood. She already had a pretty
solid feel for Pale-man. Just watching him on the integrator had told her a
lot. This is not a man who takes all the simple precautions, she said.
Page 89
ABC Amber Palm Converter, http://www.processtext.com/abcpalm.html
This is a man who makes mistakes. If he hadn t acted so nervous when we first
spotted him, if he hadn t made the slip of looking at his watch, we might have
lost him in the shuffle. Instead he brought attention to himself. If he had at
least pretended to be interested in the fight, we might have erased him from
the image trail along with everyone else who came to watch it.
And from those points you make the assumption that he would leave traces for
us to find here? Donald said.
Oh, it s no assumption, Fredda said. It s a certainty. He left something
behind. She had no logical reason for believing that, but logic was no more
than a tool of reason, and far from the only tool at that. Gut reactions had
their place as well.
Trust me, Donald, she said again, staring down at the burned-out wreck of
the security robot. This boy left a calling card.
Normalcy. The need for normalcy was painfully obvious. Caliban knew it
was so--and yet, somehow, it was difficult to act on that knowledge.
Still, the demands of the day, the strictures of routine, helped a great deal.
He had his job to do.
In theory, both Caliban and Prospero worked as field representatives for
Fredda Leving, observing the behavior and actions of New Law robots, and
reporting it to Dr. Leving s office. But their duties went far beyond those
tasks. They were roving troubleshooters, tasked to find problems that slowed
down work and resolve them.
In practice, Prospero was worse than useless in such work. He was far more
likely to urge the New Law robots to set down their tools and make for
Valhalla than he was to sort out a job-site scheduling dispute. These days,
Prospero spent most of his time with his internal hyperwave system shut off so
he would not be disturbed--or tracked. He liked to hide out from the world in
an abandoned office somewhere under the streets of Limbo, reading and writing
and studying, developing his philosophy.
Caliban, on the other hand, found that he was good at the job. He understood
at least something of both the human and robotic point of view, and could
often bring the two sides together. He had waded into the middle of any number
of disputes between humans and New Law robots--and, for that matter, between
robot and robot--struggling to find the common ground.
But there were times, be it confessed, when he wondered if New Law robots were
worthy of freedom.
For the past two weeks, Caliban had been working with a team of New Law robots
engaged in the refurbishment of an old windshifter force field coil, a
massive, powerful, and intricate device. Its repair required careful planning
and the coordination of many steps. The robot team was working without any
direct human supervision, and every robot on the team was enthusiastic about
the job.
Unfortunately, it seemed to Caliban that every New Law robot on the job had
come up with a different better idea as to how the job should be done.
There were so many ideas to sort through that it seemed unlikely that the job
itself would ever be done.
It was up to Caliban to convince the robots that better was often the enemy of
good, and that seeking perfection could mean accomplishing nothing.
It was frustrating, at times, to see the trivial uses to which the New Law
robots put their freedom. Fredda Leving had meant them to advance, to move in
new directions--not waste time around a conference table, bickering once again
over the most efficient way to retune a stasis suppression coil. He had agreed
last night to come in well ahead of schedule this morning, in hopes of
resolving a few of the issues at hand.
Come, friends, Caliban said again. Let us try again. Cannot we agree on
this very minor point?
How can you dismiss maximized efficiency as a minor point? Dextran 22
demanded.
Page 90
ABC Amber Palm Converter, http://www.processtext.com/abcpalm.html
And what good is theoretical efficiency when your enhancement routines will
leave the system unstable? Shelkcas 6 asked.
The enhancement routines are stable, Dextran replied, or at least they
would be in a properly normalized field environment.
Please! Caliban interjected. The normalization issue is resolved.
There is no need to reopen it. Friends, once again we face the same old
choice. We can solve the problem, or we can have the argument, but we cannot
do both. Dextran, your enhancement system will work, and we can use it--so
long as we do not press for greater than ninety-nine-plus percent efficiency.
Is a half-percent improvement in efficiency truly worth major reliability
degradation?
Perhaps not, Dextran admitted. Perhaps the enhancement system alone will--
[ Pobierz całość w formacie PDF ]