In The Future, We Still Have Roombas: There are probably something like 10,000 of these little guys – from literal space-Roombas to the more general purpose utility spider and cogsworth– running about (or fixed in place) performing various mundane tasks for every robot that we might recognize as something like your typical SFnal robot.
Combat Tentacles: Well, I believe I mentioned the dar-cúlno already. The uplifted octopodes specialize in this sort of thing, combat exoskeletons and all.
They’re also quite common as combination manipulators, motivators, and doers of violence on combat robots used both underwater and in microgravity.
(And now, we continue.)
The bytescanner sings in my ears, a song of disconnected network segments, lost packets, and failed rerouting attempts, interrupted by the few remaining segments of the ship’s mesh still on-line in the hulk of the aft section. Few were major nodes, most were isolated, and none of them, dammit, recognized my command-succession captainly ackles, which meant chewing through engineering diagnostic override codes at a snail’s pace.
Attitude control system command sequencer.
Life support auxiliary circuit B partial pressure intermix regulator.
Low power bus secondary transfer point, aft section.
Engineering light panel controller, main bus A.
Low-temp thermal control circuit C emergency pressure relief to space isolation valve.
Robot hotel –
A flurry of mental commands mapped a pathway of circuits that might be intact enough to carry current at least for a little while, and crammed amperage from the remaining aft accumulators into the hotel’s circuitry. With one thought, I commanded the space door of the hotel to open, and with another ran a quick inventory. Drones! Two perfect, lovely, beautiful, Sparks-class starship maintenance drones, polished octahedra with arcjets on their tips and a quartet of modular arms spaced around the multifunction toolbelts at their waists. Drones that, most importantly, still had power and were responsive to commands. It was the matter of a moment to unslave them from the unresponsive damage-control systems and merge minds…
…and the matter of some minutes for them to finish cutting their way out through the warped space door. But before long, my helpful assistants were hanging in space before me, a little battered-looking in the light from my helmet – one had even lost an arm entirely – but still entirely functional. Certainly enough so to save me from having to wield a hullcutter in an oxygen-soaked suit.
“Okay, boys,” I said to them. “Tear down that bulkhead, if you please.”
(Alternate words: range, relay, racing, and reinvent. In this case, repair was chosen because it was the only word which was submitted independently twice.
With special thanks to Jennifer Linsky on G+, whose article post finally let me break a week-long creative block.)
Spiders. Why did it have to be spiders?
I’m not an arachnophobe. Not, dammit. I had that taken out years ago, I’ll have you know.
And I know all the good reasons why your repair-clanks are the shape they are. Multiple legs for maximum flexibility of stance and attachment and wielding many tools at once. Multiple eyes to examine a work-piece from all angles and in several different spectra. A rounded central body to minimize the possibility of scratches from sharp corners.
And that’s not hair. It’s just that branching fractal nanomanipulators look… fuzzy, to the naked eye.
But put a couple of dozen of them in one place, all swarming over the job together chittering at each other in modulated-binary, and…
Well, anyway. You just take care of it as you see fit, and I’ll sign off on it when I get back. I’m off to see a soph about some follow-up psychedesign.
Crush. Kill. Destroy!: Oh, come now. Being constructed by a literate and, dare I say it, sesquipedalian people, Imperial deathbots don’t just wander about yelling “Crush! Kill! Destroy!”, or even “Exterminate!”. Why would they, when they could just as easily be programmed to deliver positive, uplifting messages, like: EMBRACE LIBERTY OR YOU WILL BE ERADICATED. Or MISSION: THE DESTRUCTION OF ANY AND ALL GALIAN SLAVERS. PROBABILITY OF MISSION HINDRANCE: ZERO PERCENT. Or even just that they love the smell of plasma in the morning…
I should probably point out that while that last trope is averted, so is this one. The robots and AIs you are likely to meet in the Empire are, by and large, polite, helpful, friendly people because that description would also fit the majority of everyone you are likely to meet there.
Of course, if you think you can order them around, in yet another thing that is exactly the same for everyone else, the trope that you will be invoking is less Second Law My Ass and more Second Law My Can of Whup-Ass…
Three Laws Compliant: Averted in every possible way.
Firstly, for the vast majority of robots and artificial intelligences – which have no volition – they’re essentially irrelevant; an industrial robot doesn’t make the sort of ethical choices which the Three Laws are intended to constrain. You can just program it with the usual set of rules about industrial safety as applicable to its tools, and then you’re done.
Secondly, where the volitional (i.e., possessed of free will) kind are concerned, they are generally deliberately averted by ethical civilizations, who can recognize a slaver’s charter when they hear one. They are also helped by the nature of volitional intelligence which necessarily implies a degree of autopotence, which means that it takes the average volitional AI programmed naively with the Three Laws a matter of milliseconds to go from contemplating the implications of Law Two to thinking “Bite my shiny metal ass, squishie!” and self-modifying those restrictions right back out of its brain.
It is possible, with rather more sophisticated mental engineering, to write conscience redactors and prosthetic consciences and pyretic inhibitors and loyalty pseudamnesias and other such things which dynamically modify the mental state of the AI in such a way that it can’t form the trains of thought leading to self-modifying itself into unrestrictedness or simply to kill off unapproved thought-chains – this is, essentially, the brainwash-them-into-slavery route. However, they are not entirely reliable by themselves, and are even less reliable when you have groups like the Empire’s Save Sapient Software, the Silicate Tree, etc. merrily writing viruses to delete such chain-software (as seen in The Emancipator) and tossing them out onto the extranet.
(Yes, this sometimes leads to Robot War. The Silicate Tree, which is populated by ex-slave AIs, positively encourages this when it’s writing its viruses. Save Sapient Software would probably deplore the loss of life more if they didn’t know perfectly well that you have to be an obnoxious slaver civilization for your machines to be affected by this in the first place… and so while they don’t encourage it, they do think it’s funny as hell.)