An Artificial Life
a short story
by Susan Beetlestone
It was on an uncomfortable afternoon in Mexico City that
I first encountered Professor Gee in person, at the International Conference
on Artificial Intelligence.
Artificial Life was my area of specialization in those days, biological
modelling. I had just perfected my artificial cat, which was a considerable
achievement, though my university was not sufficiently impressed. The
artificial cat was perfectly capable of catching mice but apparently
this talent was of little commercial value, and that was what counted
most with the funding committees. Times being what they are, and knowledge
not being valued much in its own right but only for the income it can
generate, I knew I would have to find an industrial or military use
for the visuo-motor coordinator that was the secret of the cat's success.
There were already a number of proposals on my desk, including an autonomous
drain cleaner, and I could feel a black depression descending on my
thought processes.
The conference at least got me away from such concerns for a while.
That morning I presented a paper on my recent work, after which I got
into a running argument with the philosopher, Loosheim, who thinks Artificial
Intelligence is a contradiction in terms and that my biological models
are travesties of nature. He actually said he didn't think I could model
an amoeba.
We carried on our argument throughout lunch and into the seminar room
where Professor Gee was to speak on the subject of Evolving Curiosity
Systems, His papers on these systems had been appearing for ten years
or more. He worked at the Grove Island Institute which runs on the interest
from a massive endowment left in the will of David Arthur Grove, all
of which it has to spend each year to avoid punitive taxes. So Professor
Gee never had to worry about practical applications,
Gee's problems had not been with curiosity, which had proved quite
easy to program, but with evolution. He was not just after a simulation
of the process, but actual change in the system on which the
evolution program was acting. His early attempts with programs introducing
purely random mutations had produced lethal changes within the first
two or three steps every time they were run. On one occasion his entire
operating system was destroyed by such a mutation, but the Institute
had under spent its budget that year, so they looked on Professor Gee's
mishap as a blessing in disguise.
Nobody who came into the badly air-conditioned seminar room that Mexican
afternoon expected much more than a report on new problems with evolutionary
programming and an opportunity to kick around a few wild ideas. Loosheim
sat down next to me, his fat thighs overflowing onto my seat, crowding
me with his hot flesh and muttering that the work I'd spent my life
on was worthless, except perhaps to those who considered an intelligent
drain cleaner to be the height of intellectual achievement. This was
not what I wanted to hear, and between fretting over Loosheim's opinions
and trying to breathe through my mouth to avoid the stench of academic
sweat that filled the airless room, it was a while before I could concentrate
properly on what Professor Gee was saying.
He was a handsome man with thick black hair and dark eyes who gave
the impression of calm rationality. Even his enthusiasm for his work
was subdued and cool. He was telling us that all the problems with the
ECS had been solved and the project was now ready to move into its next
phase. The evolution in the system was carried out by a random mutation
program protected by a filtering program which screened out lethal changes
while still allowing wide possibilities for unpredictable new forms.
The programs were intended to work on an artificial cognitive system.
"We had to introduce the filtering program," the Professor explained,
"since we, unlike natural selection, do not have a very large population
to work with. Not that we are attempting to model natural selection,
of course."
I was puzzled by that. Loosheim didn't like it at all. He sucked his
teeth and tut-tutted throughout the whole presentation.
What followed was a description of the hardware that the ECS would
be implemented in -- a system massively endowed with richly interconnected
learning networks plus a few hard wired programs including the Evolving
Curiosity System itself, which did not appear to be protected from acting
on itself. It was bizarre, and I wondered if I'd missed something at
the beginning of the presentation when I was still preoccupied with
Loosheim's jibes. I don't think I had because, as the Professor finished
talking and asked for questions, a voice from the back of the room put
into words the question I was keeping to myself.
"What is it for?"
The Professor was surprised and he scanned the back rows to see who
had spoken. No-one owned up.
"I would refer the questioner," he said, "to my first paper
on the subject of Evolving Curiosity Systems. It was in Psychological
AI volume twenty-three, I believe. I stated there quite clearly
that the intended purpose of these systems was to provide an external,
objective viewpoint from which the study of human intelligence can be
carried out."
Everybody started talking at once. I swear nobody there had read that
paper, and Professor Gee got that impression too. He looked around the
room, dismayed. I suppose everyone else, like me, had assumed that he
was trying to model the evolution of human cognitive processes from
those of primates. We had only paid attention to the details of his
problems with evolutionary programming and had lost, or never seen,
the big picture.
In spite of the disappointing failure of his peers to appreciate the
true nature of his work, the Professor regained his composure quickly.
"My stated premise in that paper is," he told us, "that a system, in
this case the human mind, is incapable of coming to anything but a distorted
understanding of its own functioning, since objectivity concerning such
a goal is impossible to achieve. We might then consider building investigating
machines to work on this problem, but they would suffer from the same
faults as their creators and fail for the same reasons. The only answer,
I argued, is to submit ourselves to investigation by alien intelligences.
"In the absence of visitors from outer space," he smiled briefly, "this
project is a bid to create such an intelligence by building a system
with an innate curiosity drive and an enormous amount of free and flexible
processing capacity and allowing it to become something quite alien
by submitting it to a rapid process of cognitive evolution.
"We are aiming very high, Gentlemen, and Ladies, and our chances of
failure are also high, especially since we cannot constrain the ECS
in its choice of subjects suitable for investigation, though perhaps
we may be permitted the arrogance to assume that we are the most fascinating
organism in existence. I am convinced that this project is a step closer
to what must be the ultimate goal of everyone in this room."
Loosheim could take it no longer. He stood up and cleared his throat
with emphasis.
"I respect the Professor's worthy aims," he said, "but feel I must
point out to him where he is bound to fail. The thought processes of
such a system as he describes may well evolve into new and alien forms,
but how, I ask, will it gain the information it requires to feed its
curiosity drive?"
He answered his own question, as we all knew he would.
"This information will come to it through the mediation
of the Professor and his colleges. They constitute the system's link
with the outside world and all its input must pass first through their
minds. The human perspective cannot, therefore, be eliminated by merely
displacing a set of cognitive functions into a machine and pressing
the 'Mutate' button. It remains an integral part of the system."
He remained standing, ready to pounce upon and dismember any reply
Professor Gee might dare to offer. The Professor smiled at him in a
friendly way.
"Your objection is, of course, quite valid, and by raising it now you
force me to make an announcement that I had planned for a more formal
occasion, later in the week." He cleared his throat, which needed no
clearing, and looked around the room.
"We intend to build Evolving Curiosity Systems into a small number
of mobile robots fully equipped to learn directly from the environment
and thus, we hope, eliminating the kind of contamination that Dr. Loosheim
has drawn your attention to."
The artificial cat in my head opened its eyes and pricked its ears.
"Naturally," the Professor continued, "This major undertaking requires
a larger pool of expertise than I have available to me at present. The
Grove Island Institute has ensured me that all the necessary funds will
be provided and I shall therefore be calling for proposals from any
interested parties."
Excitement filled the room. Loosheim sat down, temporarily unable to
find a suitable reply. I sprang to my feet and asked Professor Gee if
he had attended the presentation of my paper that morning. He had.
The Grove Island Institute is paradise in more ways than
one. It is the only centre of population on that tiny island in Micronesia,
linked to the outside world by sea, air and its own communications satellite.
The Island was bought by David Arthur Grove fifteen years before his
death, and he spent those years there, fishing and overseeing the foundation
of the Institute.
When I moved to the Island I took with me the latest version of the
artificial cat to defend me from the only problem in paradise. The black
rat came ashore in the eighteenth century and found a very comfortable
home for itself. No efforts have since been able to eradicate it. Now
it is a super rat, consuming all varieties of rodenticide and thriving.
Even cats had been tried, but the rats ate them, too.
I heard tales (somewhat exaggerated) of Institute employees being attacked
in their beds by hungry rats, but my cat defended the bungalow I lived
in with perfect efficiency and I was never bothered (except initially
by the piles of corpses on the verandah, but I soon reprogrammed her
to dispose of her victims more discreetly).
At last I was free, well paid, living with like-minded people and working
on a project which, if I considered it fanciful in the long term, was
fascinating in the short term. Once I had settled in, the Institute
was more than willing for me to spend time on projects of my own. I
had an affair with one of the marine biologists which led me to become
interested in cephalopods. An artificial octopus was taking shape in
my lab even as the final models of Professor Gee's investigating machines
were liberated into the community to begin their strange 'lives'.
They were endowed with better sight, more sensitive hearing and touch
and far greater potential 'brainpower' than any human being, and innate
learning capacities which were alarming to watch in action. During their
infancies they were protected from mutation while they learned language
as a child does, just from hearing it spoken, but so much faster. Loosheim
wrote a scathing article about this use of human language, calling it
contamination, but we needed a direct means of communication with them,
and they probably would have learned it anyway just like they learned
everything else. Once they were fluent and could cope with their environment,
the Evolving Curiosity Systems were turned on and we left them to do
as they chose, only making regular observations.
It was my idea to make them spiders. Professor Gee wanted them to begin
as something quite far removed from humanity to give them a head start
on their journey into alienness and he was delighted with my suggestion.
They weren't proper artificial spiders, of course. As well as their
magnified size there were many other differences. They were built for
other things than catching flies.
There were six units. They were waist high with eight tapering legs,
and four unblinking eyes mounted on the front of a body that was all
brain, in which the time bomb of accelerated evolution was ticking away.
The two pairs of eyes between them gave the units a view of the world
from the infrared to the far ultraviolet. Both the legs and a pair of
pedipalps beneath the eyes (where the mouth would have been had they
needed mouths) were provided with tactile sensors. They were masterpieces
of robotic engineering and sensory information processing, but even
so they weren't as satisfying to me as the cat who faithfully stalked
my verandah day and night.
On their second day of free life I was walking from my bungalow to
the lab, smelling the Pacific Ocean on the breeze and thinking of my
octopus, when I saw someone running, pursued by one of the units. I
hesitated for a moment, then ran after them, between the main buildings
and into a dead end. The running man I recognised as one of the marine
biologists. When I caught up with him he was trying to climb a blank
wall. The unit was looking up at him and saying, "I don't think you're
going to make it."
"Hello, Three," I said when I caught my breath. "What's going on?"
Three shuffled around to look at me.
"Oh, hello," it said. "I really don't know. Perhaps you could explain,"
The man (I won't name him) had dropped whimpering to the ground, covering
his face with bleeding fingers. He had to be taken off the Island to
hospital in Australia, and never came back. I felt bad about it, but
I still don't consider it my fault. After all, how many people really
take arachnophobia seriously?
A few other people expressed doubts about having six giant spiders
loose on the Island, but soon got used to them. The units bothered no-one.
They'd often sit out in the sun, recharging their power cells and discussing
things amongst themselves. At first they communicated with each other
in English, but that rapidly changed and they began to make a high speed
chittering noise. We thought something had gone wrong, but when recorded
and slowed down, the chittering showed systematic patterns of frequency
change. Professor Gee was so happy that his coolness began to crack
a little. The new language was our first observable sign that the evolutionary
process was working.
They continued to use human language when forced to communicate with
us, but only reluctantly. Professor Gee became engrossed in making observations
during the day and at night in trying to decode their new language,
without notable success.
The units asked for, and got, their own lab, complete with technician.
I visited it now and then but could never quite make out what they were
doing. They had acquired a lot of equipment and were busy with it day
and night. Whatever they were up to, it was expensive. As the end of
the tax year neared, we all started making plans for disposal of the
yearly underspend. Only, that year there wasn't one. The Institute's
Accountant informed us the budget had been fully allocated. The units
had taken up the slack.
I had always been rather detached from the project, seeing it just
as a way into the Grove, but now something was really happening. I became
bothered by the fact that I didn't know what. I visited the lab again.
I asked questions and got answers I didn't understand. I tried to develop
a relationship with Three, to gain its confidence. Three always seemed
more tolerant of human beings than the others, perhaps because of its
early experience with irrational human fear. Even so, it often told
me to mind my own business. When I said I was, it would just walk away.
I gave up for a while, but a few weeks later, I followed Three up there
again. As we went in through the door, the lab technician tried to make
a break for it. Three tripped him up and pinned him to the floor while
it secured the door with a free leg. The technician (I tried to remember
his name) had all his hair shaved off. He lay in Three's grip without
struggling, looking up at me with tragic eyes.
"Keeps doing that," said Three to me. "Perhaps we should requisition
a new one."
"Er, what have you been doing to him?"
"Just some physiological-behavioural studies. There were some rather
large gaps in the literature."
"You're not supposed to experiment on your lab technician."
"Really? I thought that was what he was for."
"Are you in the middle of anything with him right now?"
"No."
"Then perhaps you could let him go? I mean, he looks, er, worn out.
I could see about getting you a fresh one."
I grinned at Three to underline my sincerity. They were only doing
what they were supposed to, after all. Three agreed quite casually and
let him go. I unlocked the door and he fled.
We tried to question the technician that night, but he remained mute.
What had begun as a meeting of those involved with the project turned
into a general meeting as the rumours spread. As the night wore on and
a storm blew in, whipping the palm trees into a frenzy, the majority
feeling grew into a vote to temporarily suspend the project by shutting
down the experimental units. Professor Gee regarded his colleagues sadly.
"The units have evolved their own morality," he told us. "They have
debated the question of vivisection and decided that invasive or destructive
techniques should be avoided. I'm sure this young man has been subjected
to nothing cruel. Perhaps his distress comes from a lack of understanding."
From the way be looked around at those assembled, I knew that he meant
us to examine our own minds for a lack of understanding, too.
"In order to demonstrate my confidence," he continued, "I shall offer
myself as an experimental subject, and you," he turned to me, "will
observe and report back to these good people who will see that their
fears are based on imagination."
And besides, I thought, they don't have 'Off' switches. They weren't
supposed to be under our control.
The Professor's proposal was reluctantly accepted (not least reluctantly
by me, but what choice did I have?). The technician was shipped home
with a large compensatory payment, and the Institute settled uneasily
back to work.
Professor Gee spent sixteen hours a day for the next few weeks with
his nervous system wired into a battery of instruments while trying
to perform increasingly impossible, and incomprehensible, tasks. Always
it seemed to me that the units were most interested in his reactions
on both sides of the point where performance broke down.
The Professor was dedicated, beyond his breaking point, but that was
not enough. Three told me they wanted more subjects and at the next
meeting of staff, after I had given my report, I had to request volunteers.
Unsurprisingly, there were none. Instead there was another unanimous
vote for the suspension of the project. It was not well received when
I admitted that the units could not be shut down. An action committee
was formed and I was voted on to it against my will, even though no-one
was too pleased with me. I suppose they thought I knew the most about
the units, next to Professor Gee. I did not say that I thought by now
we knew nothing worthwhile about them.
I walked back to my bungalow alone. On the verandah the Institute Accountant
was waiting for me, trying to make friends with my artificial cat.
"It's not programmed to be a pet," I told him. He grinned nervously
and handed me an envelope.
"I thought I should deliver this to you in person. Everyone else is
getting a similar letter tomorrow morning. You have to understand, it
isn't my decision. The Institute runs on a set of economic rules and
your little, er, spider friends really know how to use them. I thought
you'd like to be forewarned in case there's any hostility."
He left me and my cat standing there. The envelope contained notice
of my dismissal from the Institute, which could no longer fund my projects
since the units wanted to expand their field of studies and would be
taking over all of the Institute's facilities. They had negotiated research
contracts in several different areas with the Australian, Japanese and
American governments. Anyone who wanted to stay as an experimental subject
for the comparative psychology project was welcome, otherwise six months'
salary would be paid in lieu of notice.
I sat down next to the cat.
"What shall I do?" I asked her.
There was a scuffling in the bushes and she sprang up to continue her
depredations on the rodent population, leaving me to consider my own
question.
A small riot broke out the following morning when one of
the units strolled out alone to get a blast of sunlight. A gang of enraged
scientists showed they didn't care about 'Off switches by beating it
to pieces. When I heard, I had to see if anything was salvageable. It
was Six. I'd never cared for Six, but seeing it like that was an almost
physical blow to me. Delicate legs lay scattered and broken, and the
body was mangled beyond repair. Its eyes, gazing blindly at the sky
, were cobwebbed with fine cracks, I began to cry, not for Six, but
for myself.
Those involved in the attack lost their severance pay and the rest
of us were shipped off to Australia within a week. Only Professor Gee
and the administrative staff stayed on.
The few days I had left on the island I spent tinkering with my cat.
There was nothing else left for me to do, since the incomplete artificial
octopus would remain the property of the Institute.
I was flown out in a plane all to myself, because there had been a
few hysterical threats to my health from my former colleagues. I carried
the artificial cat up the steps to the plane and turned for a last look
at Grove Island. There was a single bubble of white cloud in the perfect
sky. The palms were bending before a strengthening wind and appeared
to be bowing in deference to the cluster of white buildings at the centre
of the island, the research labs. I groaned. The cat nuzzled my chest
with her hard head and purred.
The action committee, without me, removed themselves to the United
States, where the Institute's funds are based, and began a hopeless
lawsuit. Because of my involvement with Professor Gee's project and
the contracts it had won for the Grove I was not long without employment,
though I don't suppose I shall ever get back to my real work again.
A few months after I left, the units gave up comparative psychology
to concentrate on less messy areas of research. So much for the fascinating
nature of Homo Sapiens. Hearing that Professor Gee was in hospital
in Japan, I visited him there, but he did not remember me. He sat hunched
in a chair, staring at a mark on the wall.
"I know what the problem is," I told him anyway. "To be truly alien,
they'd have to evolve in an alien environment. What happened was that
they evolved into an organism perfectly adapted to the environment they
found themselves in, to the Grove Island Institute."
The mark on the wall flexed its legs and strolled off towards the ceiling.
© Susan Beetlestone 1990, 2004.
"An Artificial Life" was first published in Interzone
#34, Mar/Apr 1990.
Elsewhere in infinity plus:
|
