Raindrops - episode 3 (final episode)
a short story by Mike K-H
The first episode of Raindrops is here
My screen saver was drunk. Instead of the elegant Lissajous figures the flying shapes usually executed, they were crashing into the screen boundary and staggering about before regaining their proper paths. Some instinct made me switch the sound back on, and I was treated to a chorus of the Wiffenpoofs’ Song punctuated by moans of “Ow-w-w!” whenever a shape collided with the screen boundary.
“Gentlemen songsters out on a spree -Ow-w-w-w!
Lord have mercy on such as we…. Ouch!
Ba, ba, ba-aa.”
*************
I couldn’t help laughing before I settled to the serious business of interrogating the tools I had left logging every system call since I’d reconnected to the Net.
“What on earth is going on?” Robin poked her head in, justifiably puzzled by the drunken chorus.
“Pretty much the same as in human immunology, I’d say. A small part of the colony was resistant to the ‘antibiotic’. A stimulus from outside helped to trigger re-infection, but now the colony consists entirely of the resistant strain. This is a new trick. Plenty of viruses pretend to be something else, but this is the first one to pretend to be dead.”
I put my feet up on the desk, leaned hard back in the chair, and closed my eyes. Robin waited.
“Let’s go to Tsodilo. You’ve got to go there anyway, and I need a nice long drive to give my subconscious time to do a triage of my thoughts on how to attack the problem. Semi-automatic activity like driving or washing dishes is quite an effective way of letting it get to work, and there’s more to do in the Hills than around here.”
“OK. It’ll give me longer to plan my shots, anyway.”
I had just invited myself into her workspace, an act which would normally profit neither of us since she is as unreachable as I am when she is busy. However, her schedule is partly governed by the position of the sun, whereas mine is completely arbitrary. I was betting on being able arrange to finish around the same time as she did.
I had to. Narayan had set the deadline.
* * * * *
The journey to the Tsodilo Hills was long and uneventful. Maun has had an international airport since the late nineteen-nineties, and the road around the southern end of the Delta is well-maintained blacktop.
There are still donkeys standing head-to head in the middle of the road, unmoving, in the full heat of the mid-day sun. You just have to drive round them, going on to the dirt if necessary.
There are goats, too, but they practise the pedestrian code better than most of the two-legged creatures and cross in an orderly manner - as long as some impatient idiot doesn’t blast his horn at them and make them scatter. I’ve seen more than one upside-down vehicle, usually the innocent party, caused by that kind of behaviour. Botswana is still a country where the survivors are defensive driving experts.
We set up camp by the Tsodilo Hills, and I left Robin working, getting a few background images in the evening light, while I switched my conscious mind back to my own problem. I thought about the immunologists and the way they were beginning to be watched every step of the way by moralists and religious fundamentalists as well as by scientists who were trying frantically to predict the wider effects of any new development.
The cyberworld was not without its external effects, but it was far more of a closed system, and the people who admitted to believing that some of its systems behaved so much like human beings that they must be sentient were generally living in closed systems themselves, paid for by state or private health insurance schemes.
If I let Narayan’s latest changeling turn my computer into a raving lunatic, and I sat and watched it die slowly, no-one would even try to sell the story to the tabloids.
If I created a new ‘drug’, my colleagues could test it in truly representative environments of increasing complexity. As long as they were not connected to the Net, nothing could accidentally slip through within the dimensions of the cyberworld.
There was still, however, the biggest threat to the stability of all ecological systems. Man.
However carefully we screened people and their movements, we could not be sure that someone in the team would not find a motive and a means to remove potentially dangerous code from the lab and introduce it to the real world. Perhaps there wasn’t so much difference between me and the immunologists, after all.
That thought kept me busy until three in the morning. Robin brought refreshments and refilled my coffee pot with freshly ground Mocha Java, then left me alone. She didn’t even wake when I crawled into my bag, and when I surfaced the next morning, she had already gone in search of images of the dawn. There would be birds, lizards, snakes, scorpions, dassies, and maybe even a hyaena to complement the rock paintings which were the theme of her assignment. I cooked and ate breakfast nervously, waiting for that flash of understanding that would turn last night’s work into another house of coffee-stained cards. It didn’t come. Maybe I’d got it right.
I crept into the Cruiser, where I had left the computer running all night, and inspected my patient. I’d left the machine hooked to the Net, with firewall disabled. It had been standing, chest bared, silhouetted against the skyline, inviting attack for about seven hours, and my screen saver was still producing the same beautiful geometric shapes dancing in ever-changing formation. I set off in search of Robin.
“Oh. You’ve returned to the land of the living. You look pleased with yourself, too. Have you fixed it, then?” She appeared round a pile of rocks, also looking satisfied, carrying a tripod and a large camera bag.
“I’ve done what I can here. Now it’s up to the guys at CSIR to check it out thoroughly before they release it to all the major nodes of the net. It’s going to be tough for them. That was no ordinary virus, and my fix is a first, too.”
“Can you explain in terms that I can understand?”
“Well, do you now what the term heuristic means?”
“Something to do with learning or adapting, isn’t it?”
“That’s right. My first attack on Naryan’s virus managed to identify it, and I thought it had cleaned out every trace of infection. When I connected back to the Net, I expected an identical invasion, which my code would have spotted and prevented. At first, nothing seemed to happen, then tiny bits of harmless-looking stuff were planted all over the place, at random intervals over several hours. Eventually, my monitors spotted a piece of code and waited to see what it would do when it managed to get itself executed.
“First, it collected together all the other bits that had arrive within the last few hours, and assembled a system called a neural net, and an engine to run it. This then found the pieces of the original virus which I had failed to clean out. From that evidence, it seems to have gained a few clues as to what I had failed to notice in the bits that survived, and rebuilt the virus to take advantage of this knowledge. It also added a few new mutations, giving it a good chance that something would survive the my next effort, even if I fixed my earlier omissions.”
“No wonder you were up half the night. How on earth did you produce something to fight a beast like that? It sounds worse than influenza.”
“A good parallel, except that ‘flu viruses don’t produce more than one or two serious mutations per year. Once it is out on the Net, this thing could be performing thousands of different mutations at a time - one in each computer it is attacking - and it could go on for ever. I had to play it at its own game. I’ve created a heuristic ‘antibiotic’ to fight it.
“In human immunology, I’d never even be allowed to test it, for fear that a still-imperfect version might be let loose on the world and run out of control. Even within the confines of the millions of systems linked together by the Internet, there is a significant risk that someone, somewhere, is running a system that my ‘cure’ could kill by mistake. All we can do is test it in ever-more comprehensive environments, then release it in an inactive state to each of the major nodes, with instructions on how to activate it and test it for the special characteristics of their own subnetworks. Meanwhile, we’ll ship my initial, non-heuristic fix out to try to stop the first invasion in its tracks. It should work for most people, but with several million systems out there, we’ll still find thousands who get re-infected with the mutating virus.”
“Just think. Even if your fix keeps it under control, this one new virus has added another increment to the cost of keeping every system healthy. Why do people do it?”
“Different people, different reasons. Most virus writers are pretty incompetent programmers, so it’s no good saying why don’t they use their talents in a proper job. I guess they’re all socially maladjusted to some degree - some of them seriously. Narayan is a rare case, though. His program is not state-of-the-art, but it is good solid competent stuff. “
(This is the kind of thing I was expecting from Narayan, so I’d given the concept a bit of thought already. Not so many years ago, animal and plant immunologists would respond to new forms of a pathogen by genetic manipulation of fermentation agents so that they produced a new antibiotic targeted at the new bug. Now they don’t do the redesign directly - they have developed heuristic cells which evolve new antibiotics when confronted with a new attacker, which they then use as templates in the traditional vats. I had developed a heuristic program which I had tested successfully against most of the currently known computer viruses, and I reckon Narayan may have got to know about it. He has created the first real challenge for me.)
Well, that’s it. Like many a person before me, I have been guilty of timidity in some of my forecasts (the size of hard disks on a system such as Pete’s, for a start), but some things are still to come, as far as I know. I’ll add a few tags to this post that should attract the geeks and professionals - I welcome informed comment and speculation.
Meanwhile, here are twow links - one directly connected with the subject matter, the other only vaguely so:
- The art of William Latham , whom I met while he was at IBM’s Winchester research lab, triggered the idea for Pete’s 3D image of the virus infection.
- Hyaena man . Vivienne Mackie, who contributed the Troglodyte Toour post to this blog, has a fascinating story here about a man who feeds hyaenas by hand.
Over to you. Criticise, create threads to related subjects, let’s have a bit of discussion.
