by Adam Roberts
‘But you did not give up on your ideas?’
‘No,’ said Cleonicles. ‘No, I did not. It took me a month or so, but I came up with a workable model. Now, I had to fiddle the physics a little. And the System I created looks very different from our own cosmos. To begin with, I had to make the sun burn only through nuclear fusion; there’s none of the conventional “burning” that happens in a real sun where the surface ionises and oxidises. So I wrote a sun, a little bigger than the real one, and it refused to ignite. I tried again, bigger still, and again it refused to ignite. I was left with several large globes of gaseous-state liquid. Eventually I did manage to make a burning sun, but only by accumulating so much matter – a ridiculous amount, actually, something equivalent to all the mass of our System compressed into a gaseous-state liquid ball. The sheer weight of it, compressed, the sheer gravity of so much matter broke apart atoms at the heart of the thing, and the heat began radiating. With such a large sun, I had to place the planets much further away, or they would have been cooked. So I designed an absolutely enormous System, thousands of times bigger than the real one. This solved a number of problems: for example, it didn’t fill up with gas the way ours has. I put the failed suns I’d tinkered with into orbit, and then added a few iron-cored worlds. But the trouble I had fixing the equations so that these could have an atmosphere! My first few attempts at it failed completely. Then I put in an outer world, and some moons to the larger planets, and there the temperatures were so low – so far from the sun, you see, and without the friction of orbiting in a medium such as happens in the real System – that the atmospheres simply froze. That at least gave me enough tensile strength in the material to resist sheer vacuum evaporation.’
‘Uncle . . .’ said Polystom.
‘Then I tried making a world, smaller than Kaspian-sized: I was trying to recreate my moon, although on a slightly larger scale. But most of the atmosphere bled away into the vacuum. You must understand that all through this process I was adjusting certain parameters, increasing the viscosity of the gas, sharpening the gravitational incline, that sort of thing. So not all the atmosphere vanished, but it was too thin and cold for life. I called this planet War, because it was so hard on life.’
‘Uncle, may I just . . .?’
‘I tried again, with a proper-sized world. But here the parameters were overstated, a little. I added in a mysterious “glue” force to hold the atmosphere in place. But I was too successful: the atmosphere congealed, as it were. And because this world was in a closer orbit it became far too hot – hundreds and hundreds of degrees, raging fires hot enough to melt lead. I called this world Lust, for that reason.’
Polystom put his hands together; there seemed to be no stopping and no interrupting this ghost of his uncle.
‘Finally I got it right: a similarly Kaspian-sized world to Lust, but further away from the sun, and with the “glue” force at a lesser level. Here, out of all the worlds I was writing, was one in which life could exist comfortably. This planet I named The World. Now, having learned the tweaks that nature needed to create breathable atmosphere worlds in vacuum, I could have erased the files I had written so far and started again – made a wholly new cosmos with six worlds, like ours, and put life on each. But I decided not to do that. There were two reasons for this: firstly, because my sun was so much larger than the real sun, it was necessary to put the planets considerably further away: to have the six worlds realistic distances apart, so that people could travel easily between them, would have arrayed them far too close together – collisions and mass death would have been inevitable. The only alternative was to space them much more widely, which made most of the worlds too cold. But the second reason was that, looking over what I had written, I was quite pleased with the variety of worlds, nine of them, big and small, boiling and frozen. There was that aesthetic element to the composition. I seem to remember – it’s in my memory somewhere – that you admire poetry. Perhaps you can understand the appeal of aesthetics?’
‘Oh,’ said Polystom. ‘Indeed.’
‘So I arranged my nine worlds, in their enormous vacuum orbits, around my monstrous nuclear star. I sketched out a cosmos surrounding this system on a similarly vast scale – billions upon billions by multi-billions of miles. I became quite caught up in the sublimity of the infinite, I remember. An enormous number of stars, most of them no more than a few lines of code. I’m afraid, actually, there are some inconsistencies in that – the inhabitants of the System are starting to find them out. But I had no idea they would become so advanced as to be able to determine them. The stars were really my whim, intended only as a background.’
‘Inhabitants?’
‘Oh yes, I wrote in inhabitants. Why else construct the System? This was the rationale of the whole experiment, in fact: to model populations, to discover the laws that underpinned human interaction. It’s hard to use the actual world as the basis of your observation. Our community of worlds is so stratified, so static, that human behaviour is tightly restricted. But what would it be like if several populations of millions of human beings interacted without these restraints? What patterns would emerge? The military were involved, closely, of course, from the very beginning. They hoped that the Computational Device would enable them to develop better tactics. Servant insurrection, after all, is a form of free, random human behaviour. The Counts in the military hoped that my model of such behaviour would give them insights into why some insurrections are easily crushed, and others linger on for years. The whole thing was an enormous success. Much more successful than I had ever dared hope.’
‘How so?’ Polystom asked, drawn into his dead uncle’s narrative despite himself.
‘Oh, it all functioned so smoothly! Really. I could be quite proud of myself. In the first instance the research goals were fairly large scale. I wrote in about forty fully realised characters, individuals; and then I sketched in populations of thousands and hundreds of thousands – not fully functioning, free-will individuals like the forty, but much simpler characters who obeyed this algorithm or that. I created a core civilisation, modelled closely on the antique civilisation of Kaspian in fact, located in a small geographical area of many islands clustered around a sea. About forty individuals, with complex consciousness algorithms, and the ability to reproduce these with certain variations blended genetics and environment. That was the complicated part. Then several thousand background people. And, layer on layer, I wrote in a few other civilisations. I sketched in various tribes and peoples, some here, some there. None of them were real in any sense. Then in another location – all this was within a few thousand miles of the core civilisation – I tried a little experiment. I wrote one “real” character and made the rest of the population ciphers, basic algorithms. That didn’t really work, I have to concede. The individual (I made him the ruler naturally) developed mental pathologies. It was as if he sensed that none of the people around him were real people, started believing himself the only genuine creature in the world, thought of himself as a god and so on. It meant that he regarded his population as absolutely dispensable, and acted in inappropriate ways: sending them into war after war for instance; having them give up their lives to devote to building him massive monuments and so on. But that’s a by-the-way. All the land around a certain expanse of sea – modelled in part on the Middenstead, although much sunnier – all the land around this sea was populated. Then I sketched in populations on the other continents, but these were holding patterns; there were no real people in them. And the other planets? Well, most of the other planets in this system were all lifeless, all but one. That was easy, just a few basic algorithms for physics and chemistry.’
‘And then?’
‘Then we set the Computational Device running. I wanted to see how the populations developed, how they interacted when left to their own devices. And you need to understand, my dear fellow, that in a machine so large computing time is an extremely rapid thing. We initially set the parameters at one day of
ours to one year inside the Device. So for every Kaspian-standard day that passed in the real world, a year passed for my made-up people. We watched, we observed, we analysed the data. It was fascinating: we saw whole cultures rise and fall, wars, dynastic struggles, populations fluctuate. Several of my simulated people were based on famous philosophers of our own, although, at this stage, in a very rough way. We were amazed by how responsive the person-algorithms were; advances in thought and science, although at a primitive level, were definitely made inside the simulation. They actually worked these things out for themselves.’
‘How wonderful!’ said Polystom. ‘Like an epic poem.’
‘Very much like that, indeed. It ran at that rate for six years, a little more than six years. Six and a half – of our years, you understand. In the written world of the Device thousands of years passed. Great empires rose, and fell. The focus of events shifted away from the original civilisation, the one we had originally written, which was a surprise in itself. Minor cultures grew and took centre stage, without our prompting, and the original civilisation withered and faded. We added more “real” characters, more algorithms with free will and problem-solving abilities, more “agent” characters as we called them. It was fascinating and absorbing. Then came the great change.’
‘Change?’
‘I had been involved with the programme for over eight years. It had transformed our understanding of population dynamics. Rather ironic, in fact. One of the initial reasons we set up the experiment was to help us understand servant insurrection. But we became so absorbed in the imitation world we had created that we ignored what was happening under our nose. The servants of Aelop rose up – an extraordinary, concerted uprising. Unlike anything in the history of the System. And Aelop began its transformation into the Mudworld.’
‘Was the Computational Device implicated in the uprising?’
‘Oh probably, Polystom, probably. It siphoned off considerable resources, which made the lives of the servants more unpleasant. And it dazzled us – dazzled us – so that we weren’t paying proper attention to our own estates. There was even a theory that some servants had somehow obtained access to lesser sub-systems of the Computational Device itself, and used its analytic power to coordinate their activities. Well, that’s all history now. Anyway, war broke out, and that changed everything.’
‘I was told,’ said Stom, ‘that the Computational Device was essential to the war effort.’
‘Our imaginary world, the one in the simulation, had seen more wars than you would believe. That was one of the things we discovered: without the restraints of custom and order that we are used to in the real world, humanity is an astoundingly quarrelsome race. It has an enormous appetite for war. Generals and Counts learnt new tactics from the written people of the Device: several “agent” characters had occupied positions of superiority in their armies, and they came up with a variety of brilliant novelties in the art of war. Of course, it’s possible that the insurrectionists were somehow able to access these same data.’
It was drizzling again, the surface of the pool stippled with restlessly shifting constellations of icing-drops. The rain bounced from the head and shoulders of the ghost Cleonicles, as if it were a real and material being.
‘It was decided,’ he said, ‘to change the parameters of the simulation. Our written computational world had reached undreamt-of levels of complexity. In order to be able to study it more efficiently, we decided to slow the relative passage of time inside the System. From one year passing to every day here, we changed the coding so that one and a half years passed in the System for every month that passed in our world. Three written weeks for every real day. It brought incredible levels of complexity to light, complexity that had lain buried under the onward rush of the data. And we made one other change, the most significant of all, perhaps.’
‘What was that?’
‘We introduced real people into the simulation.’
Polystom wiped the rainwater away from his face. ‘Real people?’
‘It was my idea. This is why I talk of irony, caught in my own web! When I had written the algorithms for the first “agent” characters, I had sketched them fairly roughly, or based them on famous historical figures. But they were a little crude. Many of them functioned well enough – better than well, indeed. But sometimes there were problems. Several of these characters seemed, oddly, to sense that they were more real than the people around them. There was a much higher incidence of insanity, schizophrenia, amongst the agents than we had written in for the population at large. Several of them founded religions – would you believe it? The general population was written to follow the prevailing belief systems, but these individuals (there were a couple of them) genuinely believed that they were gods, or that they talked to gods. Thus, I suppose, they explained to themselves their own sense that they were special. These new religions swept through our simulation.’
‘You were never a religious man, uncle.’
‘So I understand. It annoyed me, I know, the prevalence and variety of religious belief inside the simulation. But the decision to introduce real people into the simulation was not mine alone. It was thought that real people as “agent” characters would make the simulation more realistic, and that we would learn more from it. We took the funeral dossiers of a number of people: we collected bundles of them, or copied them, soon after the funeral oration was made. They’re wonderful raw material, funeral dossiers, particularly if they are not skimped in the writing. When a next-of-kin assembles a very detailed dossier, when the deceased has properly attended to the matter when alive, then we have almost an entire life already written. So I, and some others, transferred various of these dossiers into Computational Device coding. Dozens to begin with; then hundreds. The more detailed the dossier, the more “real” the person.’
Cleonicles stopped, looking at the rain falling around him. ‘Eventually I withdrew from the project. But I was very assiduous, throughout the rest of my life – thirty years, or more – in assembling the data for my own dossier. I kept it in a fat folder with a C on the front. And after my death, after my murder, they transferred me into the simulation. I am in one sense the most real person in it, because my dossier was the fullest. I was certainly the one with the greatest self-awareness, the greatest understanding of the artificiality of the environment. Or that’s what I thought. I thought of it as artificial. It’s . . . complex. It’s all suddenly got very complicated.’
He stopped. He seemed to be in a little pain. ‘Uncle?’ said Polystom. ‘Are you alright?’
‘It was a revolution. The introduction of real people into the System, I mean. This is about five centuries ago. Five hundred of the simulated years, I mean. A little more. Thirty of our years. We broadened the parameters of the world: certain populations that had been largely static, without agent characters, were written up in greater detail and agents introduced. And the real people acted in ways new to this world, acted in ways that the earlier, the original, agents hadn’t. They were much more inquisitive about it. As far as they knew, of course, it was the only world. None of them had the self-awareness I was just talking about. They grew up in the simulated world, and they accepted it as real, they didn’t know of anything different. Nevertheless, a high proportion of them became scientists, artists. There was an explosion of intellectual activity. Each of these algorithm-people spawned similar algorithms, that was part of the programming; agent-characters gave birth to agent-characters. And these in turn carried the restless, questing spirit into the world. One consequence was that a great many of these new agents felt the urge to explore the globe. They went on lengthy, dangerous voyages to all corners of it – we needed to revise the world with increasing frequency, to add detail and verisimilitude to parts of the globe that had been functioning, before this, on a sort of automatic pilot. So many areas of the world I had just sketched in, now agents decided they wanted to visit these places, so we had to go back and fill them in much more con
vincingly. It was hard work: inventing whole cultures, providing background detail – you see, it wasn’t enough simply to flesh out a new population in some southern hemisphere or distant latitude with agent characters of their own, although of course we did that. But we also needed to invent these cultures’ histories: to write the ruins of their past. I’m afraid we weren’t as inventive as we might have been – for example, with several of these cultures we scattered ruined pyramids in their wildernesses, really because I couldn’t think of anything else. But we were working fast.
‘It became harder and harder to ensure consistency; and hardest of all when these simulated people approached a level of technological sophistication consonant with our own. I’ll give you an example of what I mean. In my original parameters for the System, you remember I mentioned I created a planet called War?’
‘Yes.’
‘Well, it was an early piece of writing, and its atmosphere was very thin. It was a cold world, barren. But nonetheless, at that early stage, I had toyed with the idea of several inhabited worlds – it’s what I was used to, after all, in the real world. So I sketched in some basic Computational Patterns for this world. I made it mostly desert, but gave it two ice caps of water. Then I toyed with the idea of a great system of canals, by which the population brought this water from the poles to the cold deserts of the central latitudes. There was a very crude sort of culture, kings and princesses and so on: but I wrote no “agent” characters for this world, nobody with any free will. As with all such background populations, they simply bubbled along, innovating nothing. To be honest, I got so caught up with the events on The World that I more or less forgot about this other planet. Its deserts were red, and it was just visible from the World as a red dot in the sky. Anyway, long after I had left the project, one of the agent characters began examining this world, War, with the benefit of a telescope. He was a scientist, you see. He studied it, saw the canals, mapped them out, published his research. There was a great deal of interest in what he saw. It brought the distant planet into the remit of the simulation. And the writers who were working on it were faced with the necessity of inventing this entire new civilisation – of filling in all the detail in its history, and of writing agent characters for it. Who knew but that one day our simulated people might not go there? But the writers were too busy to do all this. They couldn’t really be bothered. So – do you know what they did?’