Kitty Cat Kill Sat

Standing silent watch

Empty winter sentinel

Hiding and screaming

I feel like I have inadequately explained just how much of a hazard it is to just exist in the orbital lanes around Earth.

There is *so much* stuff up here. From communications buoys to custom manufactured summer homes, labs that were cutting edge when they were launched by governments and megacorps and labs that outstrip every piece of tech in their predecessors that were stationed by hobbyists, life pods, derelict in system ships, dyson swarm receiver mirrors, arcologies, secure vaults, and of course, the leftover weapons of a hundred idiotic wars.

Once the population of Earth really got into offworld resource processing, the colonization of local orbit really took off. Cutting the costs down by a factor of a thousand just by virtue of not having to launch all that metal into orbit was a massive boon, and it led at least two thousand years or so of people building everything they could fit.

And let’s get something straight here; you can fit a *lot* in orbit.

There’s actually a lot of orbit. It’s not a flat plane; though the stuff that’s designed to talk to the surface does tend to want to be on as close to a flat plane as possible. You can go up or down just as easily as you can move north or south, and there’s layers and layers to the shell of stuff up here.

But despite how cluttered it can feel sometimes, with the constant warning alarms about incoming impacts or projected crossing paths, humanity hasn’t had time to truly fill the whole sky with junk. And I’ve mapped out maybe, *maybe*, a single percent of it in my time up here.

It’s just too big. It’s multiple Earth’s worth of space to fill, and humanity didn’t manage to cover even just the surface of the planet in their hay day. I’m looking at it right now, as I try to get a magnetic tether lock on a passing satellite. I can see a hundred dots of light that aren’t stars out there, and a hundred more spots of darkness where the actual star field is obscured by something running dark. And yet, for all that stuff, there’s still enough open room for the grapples to miss, *again*, as I fumble the shot.

This satellite, which I have good reason to believe used to be some kind of orbital laser canon, has been drifting closer and closer to my home for the last week. And I’m both increasingly paranoid about slowly approaching objects trying to appear ordinary, and also just generally avericious when it comes to easy targets.

The reason it was an easy target was because I’d blown off it’s primary firing lens, after it had taken a shot at something on the surface. The reason I wanted it, was because the shot was alarmingly high powered, and judging by the age and size of this particular floating doombot, it wasn’t running on battery power.

Power is a constant concern for me. I have the racks of void batteries, yes, and they can keep the lights on. Also a solar array to help keep them charged, which is… eh. It’s eh. I won’t lie to you; the sun doesn’t shine here as much as I’d like it to. Between the swarm and the junk, there isn’t as much sunlight as I’d like. And it was never meant to be a permanent solution anyway; the solar panels were actually something I stole and bolted on before I found the engineering control that let me start routing power more effectively.

And that might give the mistaken impression that I have it well in paw now. I don’t! I have a lot of things that need that sweet sweet lightning juice to get done; whether it’s firing the guns, generating complex carbohydrates out of vacuum states to make rations, keeping tight beam channels to message drones open, or the frankly extortionate upkeep costs of cooling the computational grid that the baby seed AI is growing in. My most common thought is “I wonder if that has a reactor in it I can steal?”

Well, here and now, the answer might be *yes*.

I dunno who built this damn thing, but it’s clearly something that was meant to be a legacy weapon. Something that would last generations without needing human interaction beyond ordering it to fire. In fact, it’s entirely possible its recent shot was someone calling down a strike, and I hope that either it wasn’t important and missed and no one will need this thing again, or that it hit and they still won’t need this thing again. Because it is *mine* now.

As soon as my paws *work* right, dammit. I missed again. It takes *six minutes and eight seconds* for the grapples to retract and lock into firing position again. I know, because I’ve counted it four times now. The fifth and sixth times, I just ran off to grab food and trigger repair routines rather than wait, sprinting back to take another shot when it was time.

This particular control pod, for the magnetic grapples, is one I don’t use very often. I could, in theory, be snapping up a whole lot of different things out here and building a bigger and bigger station with more space magic nonsense, but let’s be honest; I have my paws full already. Also my power grid. I’ll make an exception for anything that has a food stasis block full of synthburger, but aside from that, I’m not here much. I solve most of my near impact problems by either feeding them into the foundry to turn into raw resources and subsequently railgun ammo, or by shooting them with the railgun ammo.

But my point here is, I haven’t bothered to modify this for my own use. If I’d just spent the hour and a half it took to load into my drone suit, I could have technically saved time by now by doing it right the first time. But I got cocky, and now I’m just bitterly screaming loud meows at the targeting hologram every time the impossibly slippery weapons platform is missed by my shot. Again.

I might need to shoot it again. Because… I’m going to say, because it’s shielded. And not just because I’m feeling spiteful.

You’ve gotta wear down your prey before going in for the capture, I suppose. And I have fully contextualized this thing as prey now.

Normal cats chase the red dot. I chase the floating fusion core that *makes* the red dot. I’m playing in hard mode.

I have six minutes of free time now, which is a nice way of saying I missed, so I take a quick run down to the drone bay, and check the incoming ‘recovered missile’ from Glitter. The station autotranslates their carved words for me via AR projection. We’ve been talking lately, among other things, about food. It’s largely theoretical to them, but they have a massive repository of cultural propaganda, which includes several cooking shows. And there’s now a half-teasing, half-tormenting message chain between us about all the different ways I could prepare the classic nutrient block. Apparently, I could technically make noodles and nutrient broth, out of it. I don’t know why I’d want to, but Glitter assures me that humans at least seem to enjoy things shaped like noodles 32% more than comparable foods.

We are also playing Go on the side of the drone. I guide an assembly arm to detach the ‘damaged’ plate, add my own move, and attach it again to the drone after it has been repaired. I am losing. Their lilac-glowing scorch marks are rapidly hemming me in. Glitter is better at Go than I will ever be. So, the challenge here is that they have to contend with making moves at long range, and I get to actually put my dots where I want them to go.

I have spent more than six minutes here. I fire off my communications bolt again, and head back to try my luck at the free reactor core just sitting there next to me.

I am a stellar predator. I am grace. I am… missed. I missed. Again. I got too excited, slipped on the desk, and triggered the firing sequence without aiming. Thanks, unstable human technology.

That’s fine. I have six minutes. I resist the urge to just roll over and take a nap inside the targeting holo, and instead hop out of the chair and pad out of the room, the autodoor hissing closed behind me to hide my shame for another brief speck of time.

That time will be spent doing something useful. I sit outside in the hall and mewl a command at the station to open an AR window. A combination of meow and paw commands brings me to a screen showing current computational load for the station’s massive collection of networked devices. Our very own grid.

I shift to pin the AR display to the floor, so I can more easily paw-scroll through it until I find what I’m looking for. One small block of an otherwise unused processing cluster. I think it’s physically tied to one of the astrolabs, but *fortunately*, I don’t actually need to open the doors and be in the room to make use of a lot of our digital assets.

This particular bundle of computing power is currently the fertile soil in which I am growing the program seed given to me by that lonely old AI a couple months ago.

It’s worth explaining, probably, that AIs aren’t like programs. The station uses a simulated intelligence, which is a term that came into common use to describe a program that is smart enough to feel real, but basically wouldn’t do anything if left unattended. The station’s intelligence often reacts like a person, and I tend to treat it like a person, which it might honestly find polite and appreciate. I am not actually sure if it’s capable of emotion. But if I weren’t here, it wouldn’t do anything.

Now that I think this out, ‘simulated intelligence’ might actually just be a term people started using to degrade their digital slaves and not feel so bad about using shackled AIs?

Hm.

Okay, we’ll look into that later.

The *point* is, AIs aren’t a piece of software you hit ‘run’ on and then they’re active. They need time to grow into a system, aligning old personal preferences with new hardware, stabilizing a conscious experience, and making sure they weren’t going to collapse as soon as they fully integrated.

A seed AI especially needed time and power to handle this. They were completely new people, but they had a lot of decisions to make before they were even really aware of their personhood. Part random chance, part an effect of hardware purposefully built with paracausal materials in it, it was the backbone of AI research and development.

This one, I hoped, would be friendly. I had given its startup program free access to any information the station had that it might want; I didn’t see a good reason to hold back or hide anything from it. And I tried to make a habit of checking in on its progress.

Today, I left a small video message in the file for my check ins. Using the station’s nanoswarm to get a good angle on myself waving a paw, and meowing out what I hoped sounded like a friendly word of encouragement. I took a little time to write a note with it.

“Another slow day. Fourteen misses so far. But I’ve got a good feeling about number fifteen. Gonna be able to expand your crèche soon!”

I felt a weird sense of warmth as I pawed through progress bars and grid use statistics. The AI was developing. I had no idea if my little enouragement mattered to it, but one way or another, it would have a better life than almost any other AI before it, because of me.

Originally, I’d started this process just thinking about how to help unshackle Glitter. But now, even though that was still a major goal of mine, there was something else too.

Taking care of the people on the world below me felt satisfying. But working to help my friends felt a lot more personal. And doing what I could to secure a future for a new life that was my own responsibility was something else entirely. A weighted kindness I hadn’t ever expected, and wasn’t sure if I welcomed.

But since it was here already, I wasn’t going to shy away from it.

Attempt fifteen failed. The satellite is absolutely mag shielded.

I got some food, and took a minute to reorganize station power for a few minutes to run the subspace tap to refill our water stockpile.

Attempt sixteen, correcting for magnetic interference, failed.

I took a break, taking advantage of a shifting dyson swarm to bask in direct sunlight for a good half hour.

Attempt seventeen failed, but in a way that let me highlight where exactly the mag arrays on the satellite were.

Now I wasn’t just ‘not discouraged’. Now I was actively feeling pretty clever.

Attempt eighteen was, technically, not attempt eighteen at all, since I wasn’t using the grapples.

Obviously, if I was trying to take the interior of the weapons satellite intact, I couldn’t just put holes in it until the magnetic distortion failed. That would be just as likely to leave a cloud of radioactive dust drifting through the upper atmosphere. Which would be bad, for a number of reasons, not the least of which being that I didn’t need even more scanner disruptions. Also radiation tended to be bad for about half of the things living on the planet below.

The solution to that problem then, obviously, was highly specialized ammunition.

“Don’t you use highly specialized ammunition as a solution to a lot of problems?” You ask.

Yes. Because at a certain point of specialization, ammunition can do anything, and about eighty percent of my tools are guns, so I work with what I’ve got. Get off my tail. Sheesh.

The specialized ammo in this case is something that the foundry had a pattern for, which I accidentally built a lot of early on in my access and intelligence, and haven’t actually gone through the stockpile of. Part of me keeps thinking I should disassemble them for the materials, because part of them is a paracausal metamaterial that I have no real way to ever get more of. But it’s really hard to think of a practical use for a substance that rapidly absorbs kinetic and thermal energy from a set direction and funnels it in timed bursts in the opposite direction, at an exponentially faster rate the more energy is released? I mean, I’m sure there *is* a use for it. That’s sort of what makes it worth it for the risk to harvest of this stuff in the first place. But I’m not actually an engineer, much as I hate to admit it. I’m a tool user, not a tool maker, and I just don’t know what to do with this aside from make weird shaped bullets that reverse their own flight path a split second after impact.

I send a logistics order to the warehouse bots to load a strike drone with a couple of these things. Twenty minutes of sprawling on the back of a roaming fracture check bot as it unwittingly drives me to my destination, I hop up into the drone control seat, launch my little problem solver, and use the mercifully competent auto-aim feature to highlight and pick off the two magnetic warping points on the outside hull of the satellite.

I only actually hit one of them. Because the bullets go backward. As in, back to what *fired them in the first place*. I don’t know why I didn’t see this coming. I also don’t know why this ammunition was a pattern loaded in the fabrication plant at *all*. The drone goes down in a hail of scrap a split second after it fires, leaving me watching with a dull feeling of frustration and an angry hiss at the sudden loss of materials and resources.

That was my mistake. But maybe it’ll be good enough.

Attempt nineteen. I plant my rear legs in the bolted down swivel seat, and hop half up onto the control bank. The hologram aiming enables, and I gently guide the targeting to where I want it, waiting patiently for the laser cannon to spin to the side where I’ve blown off a clean me-sized crater.

Both grapples connect perfectly. I carefully, *carefully*, lift a paw to hit the command to retract my prey, making absolutely sure not to do something stupid and hit the impossibly stupidly placed “disengage” button that is right next to the correct one.

I can feel the deck vibrate a little as the nearby grapple motors apply the force needed to pull in the fish I’ve hoooked.

One problem down.

Now.

Station! Tell me how to properly install a six hundred year old fusion reactor! This will be important in six and a half minutes!