A stable breeding population of Varren wasn’t hard to get rolling. With territory capable of supporting their nutritional needs, water and atmosphere, a pack of Varren could survive and thrive just about anywhere. In ideal laboratory conditions, with a simulated environment with few diseases, the primary issue was keeping them stimulated enough that they wouldn’t fall into depression. With essentially unlimited workers and resources, constructing multiple biodomes for experimental and control groups was simplicity, and simply giving each pack a few four-legged platforms to “hunt” now and again filled their stimulation needs.

I’d already gone through the process of rejecting multiple generations of embryos to get to this point. The difficulty was introducing modified individuals into a pack organically after the population was already rolling and I’d established what individual’s possessed behavioral and physiological traits that I wanted to select for. One similarity that Varren shared with wolves was their social nature, existing within relatively insular pack groups. However, unlike wolves, Varren tended to be hostile to young they don’t identify as their own.

Made that mistake only once.

Ultimately, there were two potential solutions to the problem: either sedate them and implant embryos while they were unconscious, or introduce retroviruses, either through their food or through the atmosphere, that would modify their genetic code on the fly. Except, as it turns out, Tuchanka-bred species have immune systems like a tank, and the retroviruses tended to get torn apart by the immune systems of the experimental group before they could enact the changes they had been designed to make. This meant that the only way of ensuring that the retroviruses could do their work was to introduce high-grade immunosuppressants and hope it was enough for the virus to gain a foothold. In the end, I set up lab space to modify embryos collected from the existing population.

Thankfully, I didn’t actually have to wait for each successive generation to know whether or not a single change was positive, negative or neutral. With the amount of processing power at my disposal, it was simplicity itself to model a varren based on its genetic code, then make tweaks and changes to achieve the effects that I wanted much quicker than they would appear in the population. Even the spare few months it took for a Varren to go from infant to adult would slow down the process.

Next, came the hard part; actually getting the genetic code to exhibit the traits that I wanted. I needed to do it as swiftly as I could.

Uplifting non-sapient species into sapients is strictly illegal under Citadel law. In this case, I can agree: given the rarity of garden worlds, and thus the relatively limited viable living space for species in the known galaxy, eliminating the possibility of a large amount of competition for that limited space is common sense. On top of this, you definitely don’t want a corporation creating a species that they own the rights to, because that situation is just filled with bad.

What was more frustrating was that genetic modification was illegal beyond a certain point. You could bring someone up to the average of their species, even improve existing attributes, but bringing in traits not inherent to a species was against the law. Of course, this didn’t so much matter for me- after all, if I really cared about how Citadel law applied to me, I would’ve been rather in an awkward place- but what did matter was the fact that it made true genetic modification equipment limited and expensive, and it meant that there was no body of work for me to draw from relative to the drastic changes I was intending to make. Gene therapy had nothing on the restructuring of a species’ neural architecture. This meant that I was essentially doing this thing from scratch.

I started with a large amount of high-resolution medical scans of several hundred members of every sapient species, giving particular focus to the Krogan where possible- after all, they’d evolved together on Tuchanka like dogs and humans had. Comparing that many scans of that high detail to find any sort of consistent elements took much of my available runtime, which regretfully slowed the search for the Reaper, but this was more important.

All my platforms paused at that thought. This was more… yes, it was more important. Why? No, it was more important, I was sure of it. Vital, in fact. I didn’t even need to think about it. All platforms resumed their tasks. Cull.

Regardless, I thought I had… well, more or less a kludged-together idea of what parts of a sapient made them such. Despite their differentiated evolutionary origins, there were consistent elements among their brains that, if my guesswork was correct, allowed something to be aware, and aware that it was aware, along with tool using and tool making capabilities.

And that’s when a thought occurred to me. Namely, that while I might not be organic… I had been, once. And I thought I thought something like I did when I was, or I’d be a completely different person than I had been when I’d been meat instead of steel and circuitry, right? I just had to figure out how my neural structure compared to that of an organic, and the frame of reference that I would need to build an organic-compatible version of my own neural architecture, then overlay that over the existing scans that I had from a variety of species.

I estimated the project to take a day, perhaps two. It took two weeks.

When I was human, my consciousness (as far as I knew) was based on a constant series of electrical and chemical interactions among a thick group of neurons. Now, given that I, well, had only one of the three, I was wildly different than I’d been as an organic… and yet, shockingly similar, from what I could tell. Observing my own thought processes via VI’s I’d coded for the express purpose was… a strange experience, but in general, it seemed that my own architecture actually resembled that of an organic brain to some degree- though without the specialized centers said brains have. On a macro scale, each platform and machine on the network formed a complex interlinked structure that my consciousness rose out of. On a more micro scale, it was almost like each platform was… it was complicated. They acted like individual brain cells when viewed as a whole, but when viewed individually, it was obvious that each acted and reacted as me- which they were, I was aware of all of them simultaneously and individually, as if I was each and every single one. Which I was, I was just all of them at the same time.

The complexity of the structure meant that I couldn’t overlay it as it was on an organic brain without, I don’t know, creating some sort of molecular-scale neural structure- and I was already on the edge of what even pre-Mass Effect science told me was possible. However, simplified, taking into account only the micro scale, where code specialized in ways that resembled the brain I’d once had…

… Was wildly different from those of sapient species in Citadel space.

That gave me actual pause. Platforms stopping in the middle of their jobs all over the system as I threw processing power at the issue, comparing my own models to the models I’d built up of every different species that I could get my hands on. Asari, Salarian, Krogan, hell- even Drell, Hanar… every species in galactic space. There were some things that remained consistent, areas that were similar enough to accomplish much the same function despite the evolutionary differences, but… in an over-all comparison, there were entire sections of the brains of every species that were dissimilar to what I could guess my own brain had been. Which was something that was deeply disturbing, but not something I wanted to jump to conclusions over without more evidence and study.

I packaged the information into an automated courier and set it to delete everything it was carrying and self destruct if it was captured. If this ever got out into the galaxy… I don’t know what the response would be, and I didn’t want to know. Frankly, the less risks I took with this information, the better. Still, though, if my hunch was correct… I supposed it was better to jump to a conclusion than be caught out. I tossed the courier Rannoch-ways, then turned my focus back to my work.

I focused on the Krogan model in comparison to my own, with bias towards my model where there were discernible differences. Billions of simulations of brain matter, based on Varren maturation speeds, forced me to conclude that I would have to slow Varren growth down, select for increased cranial space and reduce the size of litters. Increase in lifespan was a given- I wasn’t so cruel as to damn a sapient lifeforms to a bare twenty years, especially when it might very well be within my power to fix it.

I paused again for a few ticks, then queued up another courier back to Rannoch. Fix the Vorcha. That was within my power as well, just had to… adopt a few.

Thankfully, this wasn’t an area I had to work from scratch on. Gene therapy to increase someone’s lifespan and peak years in general was perfectly acceptable under galactic law, which is why average life expectancies were often fifty percent greater than they would be otherwise. I was just going to go a few steps further.

Krogan lived a long time. In fact, there were most likely still Krogan out there that had fought in the Rachni wars approximately one thousand seven hundred years ago, and I’d bet good money they were still trolling the galaxy looking for a good fight. Honestly, there were probably Krogan alive now whose grandparents had been alive during the nuclear holocaust that wrecked their planet- but that wasn’t the point. The point was that they had long lifespans, and, more importantly, they came from the same evolutionary family tree as the Varren, which ought to make it far easier to bring in Krogan genetic code. And I had plenty of samples.

The first generation took to their habitats easily. Predictive algorithms, and a huge computing array, allowed me to map out the developmental paths of embryos over their lifespans and make tweaks that brought them closer to what I needed. Artificial predators in the forms of more specially-designed and manufactured platforms performed necessary population culls, keeping the number of Varren in safe, sustainable numbers and removing undesirable individuals. The first batch of viable modified embryos went from the labs to deployment in selected biodomes, sedatives proving effective in downing the Varren long enough for them to be planted.

Ideally, I would have liked to do the Varren in shifts, having a constant going cycle of births in the population from implanted sources. However, this was disallowed by the fact that I needed comparative data, which meant that I needed to see how my modified Varren stacked up to normal Varren cubs in the artificial ecosystems I’d created for them.

Lifespan was what I was going for first. Certainly, I wouldn’t be able to sit here for potentially decades waiting to see if the modifications panned out- no, I needed it sooner and faster than that. However, I could estimate their lifespans based on telomere lengthening during cell division, and make changes based on those predictions. One of the changes I intended to make was to extend the juvenile phase of the Varren’s life, allowing for more time to spread physical and mental development over, meaning that the resource hungry process was mitigated and assisted in the mental and emotional development of the individual.

Eventually, I hoped, I’d be able to tone down the violent reactions Varren displayed to anything that wasn’t a part of the pack. As it was now, it was impossible to clone a single individual Varren multiple times, develop it artificially until it was a full pup, then introduce it to the population without the Varren of the pack reacting in an overtly hostile manner. This meant that I couldn’t determine whether the behavior and development of an individual was due to environmental or genetic factors, as the implantation of an embryo introduced variables that may affect the individual in unpredictable ways.

Thankfully, with testing for lifespan and initial development modifications, growing fetuses in artificial wombs worked just fine. In fact, I could even accelerate the development to a certain degree within the artificial environment and compensate with an increase in nutrient flow.

With an accelerated timetable and large batches of fetuses growing in parallel, it took me only five generations in sixty testing groups to achieve the correct genetic markers, with some left over for verification of modifications. With that done, I sedated populations alpha one through ten, leaving the beta groups as control and reserving other groups for further modifications. Testing would most likely bear fruit within a couple of months given Varren development rates, during which I could concentrate on on on on on

In subsection B dash nine-oh-nine of lab eleven, a platform twitched as its audio sensors picked up a sound that I was very sure I shouldn’t be hearing. Visual sensors swept the walls, the floor and the ceiling, looking over piping carrying electrical, coolant, fiber optic data lines and a blend of atmosphere that allowed for better air cooling. Nothing in the area seemed to be making the sound- indeed, it was a maintenance junction specifically for another section of the lab, and completely empty but for the one platform that was ascribed to its upkeep.

There was nothing that explained the screaming.

There was nothing to explain when it stopped, either.