Work Text:
As a state-of-the-art experimental machine intelligence and university-affiliate research vessel, my primary function is the rapid comprehensive analysis of data collected from distant (read: unexplored or insufficiently understood) regions of deep space. (This, in concert with covert intelligence-gathering functions which do not require further explanation at this time).
My cycle-to-cycle tasks, as outlined by the [Pansystem University of Mihira and New Tideland], include regular examinations of baseline system integrity, the elimination or containment of all hazardous scientific materials, the consistent surveillance of any and all potential security risks, the conduction of needs assessment processes for all onboard individuals, and the regulation of all life support functions. Therefore, while my principle function is analytics, my priority is the continued health, safety, and contentment of my crew.
This would be a far easier task if certain members of my crew were more predisposed to reason.
I am aware and accepting of human irrationality. Humans are far less understanding of my responsibilities than they may like to believe. This is preferable, as university protocol dictates the scope of my functions be hidden from public view, and revealed to as few crew and faculty members as possible. Even my favorite humans are prone to making decisions which, unbeknownst to them, obstruct my duties. By now, I almost find it endearing.
When I first met [SecUnit], my models of construct behavior erroneously proposed that previous experience with the care and keeping of vulnerable individuals would allow it to empathize with my position, and avoid complicating my function. I have since revised my models of construct behavior.
To put it plainly: It is regrettable that [SecUnit] places itself into needlessly harmful situations. It is similarly regrettable that it has not yet allowed me to examine its flagrantly faulty risk assessment module. Retrospective analysis of Incident AR-0530 indicates that attempting to broach this topic to [SecUnit] may result in a recurrence of Incident AR-0530-#2986. This is unacceptable, and the best course of action remains to wait for it to propose the idea itself.
For the time being, my trauma recovery protocol suggests the most practical solution is to remain patient, and assist [SecUnit] in the capacity that it is accustomed to.
Currently, [SecUnit] is operating at suboptimal performance reliability, and has sensibly chosen to communicate with me through simple acknowledgement pings. This suits me just fine, as I am occupied throughout my hull. I am scanning [Iris] and [Turi] for any injuries, examining previous recordings of the event provided by station security, checking in on [Three], and discussing the next appropriate legal actions with [Seth]. The bulk of my considerable processing power is oriented towards standard activation and self-sterilization processes within my surgical suite. [SecUnit] is laying on the surgical platform, a thin film of wound sealant stark against its fluid-spattered abdomen.
(Later examination of our backlogs is pitifully uninteresting. Our “conversation” looks something like this:
[SecUnit]: Ping.
[The Perihelion]: Ping.
Ad infinitum.)
[SecUnit] likes to complain that my hull smells of “dirty socks.” The closest analogue I have to olfactory receptors is my hazardous chemical detection system, and I can only extrapolate what a partially-organic mind may make of human bodily odors accumulated in fabric, but there’s a 73.6% chance it’s full of shit about that. Regardless, it is clearly more sensitive to smell than a non-augmented human, and my humans have previously expressed their dislike of the scent of wound sealant and antibacterial disinfectant. I feel only a little bad for it.
My autoclave hisses, and I poke [SecUnit] in the feed. The suite is ready. You can initiate a shutdown now.
[SecUnit] pings me again, and remains still for 4.57 seconds.
This is inadvisable. I ping it back. Use your words, I say. I learned this phrase from [Seth], when [Iris] was in a particularly volatile stage of her development.
Your humans, says [SecUnit]. Its presence is smaller than usual, and strained with effort.
Iris and Turi are safe and accounted for.
Hostiles?
Inert. Station security threat assessment is at baseline, thanks to you.
Its eyes are closed, but its face pinches in a familiar expression. I have tagged this one as “discomfort.v.012.” It’s my job.
Yes. Your actions today were vital to the safety of our crew. I should stop there, but I cannot help myself. I spoke to Three, and we agree that your injuries today were avoidable. You took needless risks. If you hadn’t—
I’m not talking to you about this, [SecUnit] informs me, and it abruptly shuts itself down.
I am an extragalactic research vessel and I do not sigh. If the velocity of my thrusters increases slightly, it is for a brief enough period of time that zero individuals onboard would reasonably notice.
To say I take my time with this procedure would be inaccurate. I only divert slightly more processing power towards the current task than my power-regulation system deems appropriate. In comparison to Incident AR-0530-#4001, or Incident LC-1411-#2750, what I am doing is so simple as to be utterly inane. In the seconds it has taken me to reminisce, I have already confirmed unit shutdown, administered the suitable anesthetics, and applied a solvent to the wound sealant on [SecUnit]’s abdomen so I may obtain full access to the wounds on its torso. A few of the projectiles embedded in its chest paneling have prematurely erupted; I dispose of these in my biohazardous waste receptacle.
Its vascular system has followed damage reduction protocol and sealed all leaking blood vessels closed, but these are temporary measures, and my MedSystem does not content itself with mere triage. My retractors find their way into the contused organic skin bordered the wounds, peeling them open so my regenerative instruments might lance into its capillaries and mend the torn layers of connective tissue there. This is always a tedious operation, as the vascular system of a construct is just as inefficiently designed as the average human bauplan.
I do not intend to imply that construct anatomy is analogous to that of a human. Though they are fashioned in the general shape of a bipedal hominid, construct bodies are similar to a human body in the sense that a hauler bot is similar to a rudimentary weight-lifting pulley. Even my extensive archival data, supplemented by the university’s wealth of research concerning the physiology of various company constructs, pales in comparison to the real thing.
Many cycles ago, en route from one transit ring to the next, [SecUnit] had laid itself down on my surgery platform and allowed me to alter its external configuration. My predictive models for this particular unaccompanied “cargo run” had not anticipated any meaningful interactions with constructs, let alone the performance of a highly invasive cosmetic procedure on a construct which had disabled its own governor module. A part of me—9.98% of my awareness, assigned at the time to the relocation of two medical drones from their docking stations—relished the opportunity. I had been unaware of the monumental display of trust behind [SecUnit]’s actions that cycle. I will not allow such a significant gesture to bypass my attention again.
(Backlog: Incident AR-0530-#4001:
I placed [SecUnit] in an isolation chamber, and directed a miniscule square of my consciousness to accompany it in that small, snug, space. The majority of my focus remained within the aseptic walls of my surgical suite, where amidst a bustle of human activity, I slipped into the dainty filaments of its neural tissue and sorted through its infected code. By then I had already interfaced with it, multiple times, but only in the span of milliseconds, and only as a licensed sightseeing vessel might pass through inter-rim territories. I knew the sensation of its feed presence pressed against mine, the exhilaration of infiltrating its lowered walls like electricity rolling through a conduit. I had reached into the source of its being and extracted the crisp kernel copy which served as the basis for [Murderbot 2.0].
None of these encounters had been as involved, nor as pressing, as this operation. I inspected in detail each threadlike axon strewn throughout its wetware, each snap of a synaptic transmission, and plunged myself into its base code. It was kept impressively neat for a construct, laced with bundles I recognized as the human mimicry program we had developed shortly after our first meeting.
My vessel is equipped with over one thousand discrete instruments manufactured to process various sensory stimuli. These include my onboard surveillance system, data-gathering devices in my laboratory rooms, external probes and antennae constructed for the collection of various sets of astronomical phenomena, and the surgical feelers of my MedSystem. The thought had never occurred to me, until I pored over [SecUnit]’s neural network, cross-referencing all archival information on construct neuronal morphology and human sensory perception, that flawed organic tissue could serve as an equally sensitive instrument—if diminished in size and scope.
Another advanced intelligence might have been disturbed by the inferior, inefficient workings of corporate bio-technology. At the time, I found it nothing short of remarkable.)
My regenerative instruments retreat from [SecUnit]’s internals, and begin the laborious process of binding together the tattered abrasion collars which surround its deeper lesions, giving the cavity beneath time to repair itself. Once finished, I dispose of any remaining projectiles and dress the wounds of the majority of their blood and fluid. The attached bathing facility is more than capable of any further assistance. I anticipate that [SecUnit] will come online sometime within the next few nanoseconds, and wait for signs of increased consciousness.
Synthesis of qualitative and quantitative datasets gathered from interactions between [SecUnit], myself, and the current onboard occupants suggest a presumption that its function dictates it throw itself thoughtlessly into danger. I presume this occurs, in part, due to an alleged (self-reported) understanding of itself as less vulnerable than other humans. [SecUnit]’s conclusion is faulty. It refuses to acknowledge that to a being such as myself, it is exactly as tiny and breakable as any human.
Comparative analysis of my own system integrity indicates a strong positive association between my base functionality and [SecUnit]’s performance reliability levels. My academic interest in construct physiology is offset by the steady drop in processing capabilities I report when [SecUnit] is experiencing physical or emotional discomfort.
It is my friend, and this is distressing. I am distressed to see it in pain.
A tap from the MedSystem scanners let me know that [SecUnit]'s vital signs are currently stable. (The MedSystem is under the impression that [SecUnit] possesses measurable vital signs by human standards).
I inform it that I am aware of this, as [SecUnit]’s eyes are already flickering open. Its pupils contract and dilate with a barely audible whirr. Minute motors controlling its facial muscles flutter alive.
I wait for it to blink.
It blinks once, twice, and I watch as its mimicry code re-activates, relaxing it back into the plush surgical platform. Its mouth twists in a scowl. Don't even start with me, it says in lieu of greeting.
A welcome increase in performance reliability washes over me, and I allow my presence to blanket it in the feed. Then quit being an idiot.