Lovelock (25 page)

I lay there, panting. On the bed above me, the bouncing grew still; they, too, rested, breathing heavily.

I will have to clean up this spot on the floor, I told myself.

But then, after this slave thought, another idea came to me, swept over me. I am a free man now.

I must have made the decision unconsciously long before that night when I lay under Carol Jeanne’s bed and listened to her and Red trying to make a baby. Because when they were both breathing slowly and heavily in sleep and I slipped out from under the bed, there were things I didn’t have to look up on the network. I didn’t have to look up whether there would be time enough for one of the capuchin embryos to gestate before launch, because I already knew. I didn’t have to look up where the gestation chambers were, nor did I have to check to see whether they were being guarded or checked in any way. I had already looked up all this information. Accidentally stumbled across it, actually, on the way to looking up other things. But I remembered it. Why did I remember it? Because I must have known that somewhere along the line, I would overcome the inhibitions on sexual pleasure they had forced on me, and I would be able to mate.

Yet perhaps I didn’t know anything of the sort. Perhaps I simply desired it so much that I had to find out those things even though I never believed, at any level, that I would win my freedom from even one of the shackles they had placed upon me. Revolutions begin, not with foreknowledge of victory, but with such deep and powerful desire that the question of success is not a part of the equation. The attempt will be made, no matter the odds, no matter the utter lack of rational hope.

I still had two problems to solve before I hatched me a monkey. The first was easy enough—to alter the embryo inventory to show that one more of the capuchin embryos had been among the transfer wastage than the current records showed. I couldn’t just change it on the current records, of course. I had to break in to the secure backups, which required writing a little bit of temporary piggyback code on the backup software that would make the backups agree with my inventory without reporting the discrepancy. Not hard.

The harder one was this: The new network software should have been online the week before, and I couldn’t count on it being delayed much longer. When it came on, my access to the system would probably be severely limited. With the old network, thanks to what I had learned while tracking Peter’s little message, I had learned how to navigate pretty easily. But if I was going to be able to maintain secrecy, I had to have as much power in the new software as I had in the old. And it was deeply unlikely that the new network operating system would have convenient back doors left in by careless programmers. Those days were long gone.

The thought did occur to me once that maybe I didn’t need to conceal what I was doing. Maybe I could just tell Carol Jeanne that I had overcome my programming and I could probably hump some nice little piece of tail (why do
humans
use that expression?) if she would just do me the favor of thawing out a little capuchin bimbette for me.

Naturally, Carol Jeanne, being a loyal friend with such deep respect for my rights as a person, would go straight to the security people and tell them that her witness’s programming had failed and dear little Lovelock would need to be destroyed. The toaster is broken, and I’ll need a new toaster…. Oh, too bad, there
aren’t
any more toasters? Well, I’ll just make do without.
This
one just isn’t safe anymore. Maybe you can cannibalize it for parts.

No, I’d keep it a secret, thanks.

I needed every spare minute to explore the new software, devise my control system, program it, and install it undetected by the sysops. Yet, officially, witnesses didn’t have free time. I was supposed to be with Carol Jeanne every waking moment of her life, because humans have the idea that every last thought or action of a famous human is cause for admiration and speculation, of great import and interest to the teeming masses.

Fortunately, Carol Jeanne was well aware that humans would be venerating my biographical dumps on her for generations, and she was ill at ease with the idea of college students studying her bathroom habits or her procreative efforts or her extramarital flirtations. Just as in other systems of slavery, there were times when this master didn’t want her servant around, and the servant could
pretend
that these moments of disregard were “freedom.” Thus I had free time.

Lots of it. Carol Jeanne’s little experiment in babymaking didn’t change the fact that Red was being an unbearable mama’s boy, and Stef’s message had done its job—Carol Jeanne wasn’t accusing anybody, but she was avoiding Liz. Neeraj, however, was still there, still charming, still genuinely appreciative of Carol Jeanne as a scientist, as an administrator, and as a woman. I don’t think Carol Jeanne and Neeraj were making the beast with two backs yet, but they were enjoying some heart-to-heart conversations and extended work sessions, and more and more both I and the cockatoo were not needed to witness such “boring” and “routine” work when we could be doing “vital” assignments in filing and research.

The research I spent my time on was the region of the network where the new software was being designed, and it was not easy to breach. The sysops might not know the back door that Peter and I had found, but they certainly knew the old system leaked, and so the work on the new system was taking place off network. All the computers working on its design were disconnected from the rest of the computers in the Ark. For a day I despaired.

But human beings aren’t perfect, right? They even take pride in this. “I’m only human.” They say that a lot, especially when they screw up and want to be congratulated for it. So I found ways. A lot of these programmers took work home. They had to handcarry tiny little disks, and they were all very conscientious about erasing their work when they signed off. But it wasn’t hard to install little routines on their home computers—which
were
tied to the old network—that would make clandestine copies of everything that they erased. This close to the end of the project, many of them were debugging high-level interoperativity problems, which meant they had to install much of the finished software in order to test any part of it—and all parts were being tested. In three days I had assembled, in bits and pieces, a library consisting of, as near as I could tell, the entire network system.

How to get into it? How to hide? I could install a back door, of course, but it would be hard to make it unfindable. I had found the back door to the old software just by checking the routines that read keystrokes. If anyone started getting suspicious of me, that kind of back door would be easy to find.

I studied the plan of it, how the software worked, how it kept unauthorized users out, how it checked its own integrity. File sizes and parity checks were continuous; I couldn’t alter the code. Once it was running, I couldn’t access the underlying systems without leaving tracks.

So what I finally did was this: I wrote a little program that lived in volatile memory all over the existing network. It functioned only during hardware interrupts, and it hid its memory use and storage in unused disk space without telling the operating system it was there. If it was about to be overwritten in one place, it moved itself to another. It evaded all the software they used to check system activity. And when they made the benchmarks against which they would measure future system performance, my little program would already be there, so that from then on, “normal operation” would include whatever processor cycles it stole.

Besides hiding, what does my little sleeper program do? Why, nothing that anyone would notice. It replicates itself at every opportunity so it can’t be erased. And it checks keyboards for my own little entry code. Which I’m not going to write here because I don’t know for sure that this file can’t be found.

When it detects my entry code, my sleeper does something very simple. It allows me to replace sections of the operating system with my own altered versions. I can make new ones at any time, on any computer, and slip them into place. As long as I program well and don’t crash the system, I can replace any section of code that I want, and while I’m using it my little sleeper program will protect it from all error checking. Then, when I’ve done what I needed to do, my sleeper puts back the original network code and my special version goes back into hiding in secret unfindable places on disks scattered throughout the Ark.

Unfindable? Well, actually, nothing is unfindable. But it’s very hard to find, and my sleeper watches to see whether someone seems to be looking for it. If they are, my sleeper destroys all copies of my programs hiding on that particular disk. It won’t matter—there’s always another copy somewhere. And if by some miserable stroke of ill fortune they manage to find and destroy every copy of all my programs on disk, my sleeper will still be there, ready to let me in to write new ones. Because they can’t ever, ever get rid of my sleeper. Not unless they shut down every computer on the Ark. And if they did that, the Ark’s life support systems would cease and everyone would die before the computers could get back online.

I thought my solution was simple and elegant. It would work. Until the new network was put online, I would use the same back door that Peter was using. After that, I would be the only one with special access.

It took me five days. I’m really very good at this. After all, I’ve been enhanced.

The gestation chambers are completely sealed off until we get to our destination. Then they’ll go like gangbusters for a few years until the new environment is stabilized, after which they’ll be useless again. Everything depends on them, since this is where Earth species of edible, employable, or ecologically necessary animals will be revived from eggs and frozen embryos. There has to be a lot of space because literally thousands of animals will be needed at once.

I only needed one. And after all my planning, it was really pretty easy. I removed the single female capuchin monkey embryo from one of the icehouses, took it to the most remote gestation chamber, and got it started.

Of course there was more to it than that. There was the computer work: altering the inventory lists, making sure the backup software didn’t catch the discrepancy, and then rewriting the gestation chamber monitor software so that it didn’t report on the one operating chamber and yet still allowed it to run. That was the most complicated part of the task, but once my alterations were in and running, there was nothing more to it. Then it was a bit harrowing, moving through the air circulation system and crawl spaces, finding the right icehouse—one of forty square tubes, three meters on a side, which were always kept at —40°C—and then climbing down when nobody was looking to get the embryo. It was intense.