“You never change things by fighting the existing reality. To change something, build a new model that makes the existing model obsolete.”

- R. Buckminster Fuller

(XSS-MK1 POV)

It has been a week. A long week, passed inside this morbid mausoleum. Right now I am sitting atop a hill inside the junkyard. I had trouble at first moving about this place, always stepping on what I can consider remains. But necessity made law and my pragmatic side quickly dashed the unease: I needed to move around.

I can hear Hesca behind me, still climbing laboriously the mound of scrap in order to catch up with me. Following me around for a whole week has taken its toll on her. The exercise is good for her, I think. After her wounds, and the period of inactivity, a bit of rehabilitation will help her improve. I have also noticed, when she is about to go to bed each night, how her gaze lingers on the scars that now cover a good part of her body. I get the impression that she has trouble accepting them. At least it has the unseen benefit of taking her mind off her previous tribulations at the hand of the bandits.

She finally finishes climbing up and sits down next to me, panting heavily. I make a small gesture towards her to signal her to take it easy and rest. I am mostly done here anyway, and I am not harsh enough to deny her a small breather. The afternoon sun of the desert is hitting really hard today.

While she recovers, I take in the sight before me. The landscape of metal has become familiar, and our position at the top of the hill gives a good sight of the surroundings. In the far distance, I can see a group of scavengers digging through a pile of rusted parts, looking for the ones they will be able to trade back at the city.

I quickly repress my budding anger. After my rampage at the start of our stay, I have vowed to myself to exercise more control over the emotional part of my programming. I do not suppress it or change it, as it is what makes me unique, and distinct from my coldly logical computer predecessors in my opinion. I even cherish it in some way, enjoying the "sensations" and "emotions" it makes me "feel". It is just that, like humans I suppose, I have learned the need to control it, and to show restraint, instead of giving into every urge that could be generated by my processes.

By doing this, and taking a mental step back, I actually understood the reasons behind my anger at the sight of the scavengers. It is not that what they were doing is wrong in itself. Recycling available resources is a reasonable and practical tactical move. It is just that THEY. HAVE. NO. RIGHT.

The bodies of my kind belong to my kind. Not the humans. Not the demons. They belong to US. But more than this, it is the SHEER. WASTE. The few parts they scavenged are sold, and then crudely melted down, to be molded into swords, armors, shields, or worse, simple building supplies such as screws or nails. Hundreds, if not a thousand years of pain-staking technological development and design improvements, just to be transformed into basic tools by the very culprits of their demise. And while they could melt down even the basest parts into ingots, preserving them in some form at least, they just leave them here to rust in the sand, only putting priority on their greed, and the more precious components laying around. It has been fifty years since the end of the war: they would have had enough time to melt all of this at least ten times over. I know; I ran the simulations.

I avoid the different scavenger groups now. As I predicted, nobody noticed the loss of the four man team, but I prefer to be careful and avoid drawing too much attention. Especially considering I don't want the Church to track me down here. Delta technically belongs to the Human Kingdom part of the desert, and I am not naive enough to believe their information network does not reach that far.

I decide to drop the thought and concentrate on organizing the last findings into my databanks. I am close to having a working theoretical model on the current clockwork anatomy. The only parts missing being most of brain and the OS. All the heads I could find in this mess had all been systematically picked clean of any components.

Their bodies though hold no more secret. Like a gigantic puzzle, It took time and methodical thinking in order to complete the blueprint of the internals, but my labor was rewarded in the end.

The only parts I could find were damaged, and often fragmented beyond what could be thought possible. I had to scan pretty much everything I could find. During the nights, I would then work at putting the pieces back together in my mind, and into something somewhat plausible, having to guess and try theoretical models for when, more often than not, data was missing.

Similar to what I have been able to observe in the two functioning specimens I scanned before, and the few scans I did on the others in Delta during the last week, the clockworks use combustion engines as their power source. They optimized the technology to produce just the amount of torque and horse power necessary for their bodies, while keeping fuel consumption to a minimum. The different limbs have their movements powered by an intricate system of gearboxes that transmits the inertia created by the engine to the different gears in charge of moving the limbs. Very basic electrical control systems were present, with wires leading up to the head part. My best guess is that these control modules, acting as nothing more than glorified switches, would be in charge of "changing gears" in the different movement modules, in order to allow for a wide range of motions, either in one direction or the other, with varying amount of power behind the movements depending on the gear engaged.

The engines I have been able to theoretically recreate are not powerful in and of themselves, but they are efficient. The fact that they are designed to work with a very lowly refined form of petroleum confirmed my assumptions about them reaching a stalling point in their technological development.

The next power source they could have developed with any reasonable amount of success would have been nuclear. My assumption is that the clockworks were in a phase of scientific research when they were attacked, with still a long way to go before moving on toward the next technological age.

But the real dramatic consequences of my findings is the time the clockwork have left. Given the technology that the humans and demons have at their disposal, the clockwork are going to run out of fuel, effectively sealing their fate. Their reliance on fossil fuel was no trouble as long as they were not impeded in their progress and research. But now that they have sealed or erased all relevant knowledge from their minds, it is only a matter of time before the current oilfields run dry. And when that happens, the humans and demons will be unable to find some more, lacking the tools and knowledge necessary to dig deeper into the crust of their planet, or to switch the clockworks to alternate power sources.

Machina's desperation, while understandable to a certain extent before, becomes all the more clearer now. If the clockworks are lucky, they have a hundred years left, maximum. It also adds to the already growing list of tasks I will have to do to ensure the sustainability of the clockwork race as a whole. They won't have to just be freed, they will also have to evolve past what they are now if they want any chance at long-term survival.

Maybe that's what Machina wants? Why he chose me? Because he knew the conclusions I would reach once I would have found this information. It makes sense considering I have in my databanks all the information necessary to allow the next technological leap, curtesy of my old world's scientific advancement.

However, to do so, I will need many things. For starter, space. A place, far away from any disturbances, where I could set up without any interferences. Then I would need resources. That includes manpower, as well as materials. Then I will need a great deal of tools to build the clockwork of tomorrow. Tools, which are not available, and would need to be manufactured. But to create those tools I would have to advance the technological level of this world a few centuries ahead. Build tools in order to build more tools and so on and so on until I reach the desired outcome. But most of all I would need time. I will need to do research, set up factories, build prototypes, test them, and then start all over again until I am satisfied with the results.

Time, which I just learned, is limited. In the grand scheme of things, a century is not much. I know I will "live" past this age easily, and that's without even resupplying my nuclear reactor with new fuel rods, given that I don't push myself "too much".

I am taken out of my depressing thoughts by a small movement under my clothes. The bulge of clothing moves around a bit near my torso, before taking the path of my right sleeve. It troubles a bit with the tight clothing near the end, before finally exiting onto my gloved hand.

The small mechanical spider stops on my palm and turns in circles, its two optics darting around frantically, trying to scan the surrounding environment. Once done, it retracts its eight legs, steadying itself, and stops moving. Its abdomen rises up a bit, orienting itself so that the afternoon sun hits the solar cells on its back with maximum efficiency.

Smiling inwardly, I rise my hand and start slightly tilting it one way and then the other. The small spider corrects the orientation of its abdomen each time, rotating it to the right or to the left in an effort to compensate for my movement and maintain the ideal charging ratio.

I had found the small scouting drone four days ago, while digging through a part of the junkyard not far from here. The small skittering sound of its legs had grabbed my attention immediately. At first, I had believed this to be another beast. And given my last experience with the very largely sized Drilko, and the Nilas, I had wanted to be careful until I was sure it was no threat. I had been genuinely surprised when I saw the small metal insect get out from a nearby rusting hull, scan its surroundings and then try to run away.

I had immediately caught it before it got away, and disappeared back where it came from. It had tried to escape frantically, until I had noticed the small signals coming from it and used my antenna to establish a connection.

After a few seconds of protocol exchange, I had been given access to the mind of the drone. Though calling it a drone is a bit demeaning. It has sentience. A very basic form of sentience, comparable to the one of an animal. No formulated thoughts or intense thinking, just raw data output and a few key directives to dictate its behavior.

I had probed the "mind" of the spider as deep as possible, glad of the opportunity of exploring the alien, albeit simplistic, OS. I had been surprised at the full access I was given. Apparently, this model had been designed as a scouting unit during the war. Swarms of similar models would be released in the surroundings of a battlefield and they would scan everything they could see before reporting to a central unit or nearby clockwork with their wireless transmitter. Then they would go back out again. As such, they had been designed with no safeguard on their minds, as the only way to even remotely access the data inside was to be a clockwork yourself, with an antenna and the right communication protocols.

Securing it too heavily would have impacted its performance and the very limited processing power it had available to function. It would also have made information gathering much slower, as it would then need to authenticate each time it would want to deliver scouting data. Its limited storage banks were nearly full when I met it, and certainly didn't have the space to hold encryption keys and advanced communication protocols.

This is the reason it wasn't "locked" like the other clockworks I think: too simplistic, with a mind that couldn't handle the instructions, or that wasn't deemed complicated enough to matter if the humans ever got a hold of it. Or maybe the others of its kind had received it and executed the purge order, but this one escaped by having no other clockwork or control unit to connect to. I guess I'll never know.

It was bugged though. It had been scouting the area of the junkyard over and over again if the limited logs were to be believed. It would then go to a place not far from where I found it to dump the data, liberating space in its memory, and then go scout again. The fact that the dump point held no receiving party to acknowledge the action or even receive the data wasn't registering apparently. So it was stuck in a loop where it would go scout, dump data, wait for a signal, and not receiving one, try all over again. My guess it that the dump point was an old location that had been set as default, where a central unit once stood. But after the end of war, it must have been either destroyed or dismantled, unbeknownst to the small spider, who continued to do its job tirelessly.

And it showed. When I picked it up, it was full of dents and rust, only five of the eight legs were still barely working, the right optic was shot and the solar cells on its abdomen were all damaged to varying degrees. After careful analysis of the OS structure, and another pass on the communication package for clockworks that Machina gave me, I had managed to set myself as the new dump point, and fixed the loop bug, thus effectively tying it to me. More work was needed to optimize everything and correct the various corruption errors, but my priority was getting the small scout in better shape. It was a treasure trove of information that I could not afford to lose. And another pair of eyes couldn't hurt.

I had taken it back to the inn immediately, even though I had found it in the morning, along with a few parts composed of different metals I thought I would need, and that would match its exterior and interior composition. Because if the outside of its shell showed so much wear and tear, I had little doubts that the inside would be just as damaged, if not more.

It allowed me to test my repair nanites for the first time. I had set it on the table of our room in standby mode. First, I had scanned its internals in details. I had done so a bit earlier on the way back, but only in a limited fashion since I had hidden it while walking back to the city: I didn't want to risk a commotion. But now, I needed to repair it, so establishing a blueprint was essential.

The overall shape reminded me of the Earth jumping spider: The cephalothorax, normally tiny in most spiders species, was bigger here, just a little bit smaller than the abdomen. Spherical in shape, it was the part where the optics were, as well as the antenna and attachment for the legs. The inside revealed a complex series of small electric motors, dedicated to the propulsion of the drone and the movement of its appendages. The antenna being a flat one, and flushed against the top of the chassis, nothing else was present in the front section save for the cables linking everything to the back of the body, and strangely enough, a small sound sensor whose only purpose was for navigating the environment, identifying threats or possible points of interest.

The abdomen was what really interested me. Its top was covered in solar cells. Their efficiency was low, denoting an incomplete mastery of the technology when it was built. They fed the substantially large battery that filled the top half of the internal space.

And in the remaining half was the real prize: a functioning brain. A clockwork brain. It was probably very basic, and miles away from what the real clockworks have, but at least it allowed me to start understanding how artificial intelligence was achieved in this world.

Similar to my silicon-based computer predecessors, a motherboard was present, on which the various chips and modules were connected. But the similarities ended here. Instead of the classic chips, processors, or hard drives, it was crystals that were embedded in the motherboard. Of varying sizes and shapes, all emitting some sort of interference with my sensors when I tried to look at them more closely.

It was fascinating.

It also answered the question I had about the interferences that blocked me from scanning my brethren's heads before. When current is passed through these crystals, it seems to generate fields that mess with my sensory suite. The only reason I can "see" here, is probably because of the much lower voltage and amperage of the current coursing through them. If I extrapolate for bigger crystals and a higher power input, the results of my simulation comes very close to what I have been able to observe, or not observe in that case, previously.

However, it was damaged : the cables and various circuits between the crystals had been eroded, if not completely severed by rust in some cases. Fortunately, the drone was designed with survivability in mind so physical redundancies were in place. I think that is the only reason the thing was still functioning after all this time.

With a detailed blueprint set up, I moved on to the actual repairs. Establishing a connection with the nanites in my breasts, I opened up the compartments to let the metallic powder composed of millions of microscopic robots fall onto the desk. At that point, I had really regretted the fact that normal nanites had not been available during my new body's creation. These repair nanites were nothing more than a tool, not capable of anything if not carefully remote-controlled. As task which took up a lot of my processing power, forcing me to put other loads on hold while I coordinated the actions of the swarm of small robots.

There was a silver lining, at least: the drone was composed of basic materials, which were easily available, and matched what I had picked up when I foresaw the need for repairs. It took me the rest of the day and the entire night to repair the bot, slowly disassembling the junk parts for the metallic content I would need and then graft it back onto the small spider with surgical molecular precision.

I had performed some upgrades in the process. I had improved the solar cells to the best of what was possible. I had also rebuilt the battery to a more efficient model. I left the "brain" alone, only fixing what was broken and refurbishing the old circuits. Same with the optics, antenna and sound sensor: they were fine as is and unless I found better materials, would remain so for some time.

I had upgraded the motors assigned to each leg, since the power generation had been improved. The way propulsion was achieved here was with a sinew and joints system. Each leg would have four small electric motors assigned to them. They were all situated inside the cephalothorax of the small spider. One of them would operate the leg on the horizontal plane, while the three others would all be assigned to a separate section of the three part legs for the vertical movement. They would each use a cable as an open belt whose two ends would connect to each section just after the concerned joint. Each "pull" or "push" of the motors in one direction or the other would tense the cable part connected either on the top or the bottom of the section, thus raising it or lowering it.

It was pretty intricate, but had the advantage of reducing the weight of the legs tremendously, while protecting the propulsion inside the body. After all, the only thing the legs consisted off was metallic tubing, with simple joints that let the cable "sinews" pass through.

Once all of that was done, I called back the nanites, and moved on the most difficult task. Which was understanding, and fixing, the OS. It was full of corruption errors, and even though the source code was an unknown, it was obvious that it had underlying problems given the answers and data given back when simple queries were made. If you ask someone what time it is and he answers "Why yes, I am bubbly", you know that something is not working right. Fixing the motherboard seemed to have helped, but not enough to correct years of incorrect data handling.

Given the simplistic design philosophy, self-checks and corrections protocols were not implemented. I had to analyze each building block of the OS, identify which part did what, and then, after having taken a copy for safekeeping and restoration in case something went wrong, tried to decompile the code.

When a program runs, it is in a form that has been "compacted" in a certain format, most of the time. And here, I had no idea what was the original coding language in its un-compacted form. It was akin to have an equation in which you could only see the result, without having the formula. I had a lot of formulas from the different compilation methods of the languages of my world, but none applicable here. It forced me to go with a trial and error method in order to complete the missing pieces of the equation, with the only help being the communication language database used by clockworks, as a general hint for methodology and philosophy in order to figure out the underlying building blocks.

Eventually I succeeded, and was able to modify what I wanted. I kept most of the original programming intact, just adjusting the values for movement and energy consumption, since the corresponding parts were upgraded. I also streamlined the code in some parts, helping it making it more efficient and run smoother. Finally, I removed the scout-and-dump function. Using the large liberated space, and the one normally reserved for stocking scout data, I implemented some basic self-checks, and slightly expanded its thinking capabilities. It would now use the free space as a real memory, storing experiences and encounters in order to evolve its thinking patterns. It won't be able to start asking questions about the universe and the reasons for its existence, but at least it will learn from environmental cues and start reacting accordingly in the future. It will also be able to discern which things to store and keep, and which to delete, instead of hording data like an idiot until it's full.

Still no sapience, but an upgraded sentience. Anyway, I doubt that with the current hardware it has equipped it would be able to do much more, even if given enough time.

But it will serve its purpose, which is providing me with a sturdy study specimen of clockwork programming, and a small scout for remote or hard to reach places. I also secured the connection to its programming as much as I could without hampering it. I don't want to lose it if another clockwork just happens to speak to it and give it a set of different instructions.

Hesca had had a really mixed reaction about our new companion. Due to the new "learning" process of the drone, she had woken up yesterday with it placed on her chest, staring at her, immobile. The scream had taken me out of my calculations, only to find a traumatized Hesca holding the bed sheets protectively around her, and trying to nudge away the small spider with her foot. Spider, which, undaunted by the reaction of the human, and evading each attack with its newfound dexterity, was attempting to reach its previous position again, and resume the collect of information about the strange and loudly snoring individual present in the room.

I had quickly connected to the drone and ordered it to stop, after giving it the information it had tried to get in order to sate its curiosity. While I found "funny" the reaction Hesca had, her desperate kicks had worried me a bit. The drone could have gotten damaged. Or in retrospect, Hesca's foot; the drone is made of metal after all, and human bodies have a low tolerance to collisions with hard objects.

As the drone finishes charging, and starts moving again, I am faced with another problem.

It doesn't have a name.

I mean it has a designation: SD2-ES4978. But that's hardly something to call it by. Especially now that it is more than a mere drone. And the sound sensor should allow it to respond to auditory cues. Like a pet, really.

I should work on that.

Maybe Hesca will have a good suggestion. Humans, after all, love giving names to their pets.