Well that was a difficult choice. I liked all of those approaches, I already used them regularly.
I ruled out precision machinery first. While it would in theory help with one of the more pressing problems with manufacturing that I currently faced—that I had a lot of potential manufacturing capability but not a lot of resources—it wouldn't actually help that much as my current manufacturing loss was less than 2%, and I was perfectly capable of recycling that lost material. The second benefit of stretched tolerances would require me to have both access to more advanced object designs and worse tool designs than I currently did.
Modular production lines would allow me to quickly adapt to new situations and fit my preferred strategy of constantly improving. It would be worse in the long term as being able to swap my production on demand could only offer me marginal throughput increases—and only when I regularly changed what I was manufacturing.
Specialized tooling would allow me to out manufacture any other factory, I had room to spare so I could in theory build plenty of specialized lines—which at the moment tended to average closer to an 80% improvement, this perk would provide an effective +20% to that (though technically, in some cases, it would provide no benefit at all). Simple and straightforward.
In the end however I decided on Modular Production Lines. I was currently limited in working machinery and I had a lot of different things to make with what were effectively scraps. This would require me to constantly change up my manufacturing processes on both the input and output end—and hence everything in between.
While it was tempting to be able to outproduce anything else, it wasn't currently a mediating goal. Besides the brute force that came from being a city size factory would compensate for that.
Every rank apparently gave an additional 2% bonus to throughput for all manufacturing—the "manufacturing rank bonus" currently sitting at 4%. It should be 12% in a moment once I was at rank 6. An applicable Manufacturing Focus would then in theory double that to 24%.
Throughput was a strange property to increase from my point of view. The equation for throughput (all other things being held constant) implied that the bonus would allow me to either change the speed of manufacturing a single item, increase the amount of items I could manufacture on a given production line simultaneously, reduce the amount of space I needed for a production line, or any combination of those that worked out to match the bonus. But those equations had deeper variables. I could potentially find some other effects. Something to experiment with later.
There were a number of keywords I could pick from for the specialty apparently:
agriculture-machinery, agriculture-materials, airships, and so on. Some more focused than others, and a lot of them overlapped. The implication of overlaps did not escape my attention.
Simulations of choosing Manufacturing Focuses had found two competing clusters of promising heuristic candidates. One was focusing on my maintenance mechbots, manufacturing machinery, computational elements, and so on, as this would increase my ability to improve and repair my self-as-factory. Improving myself was core to my being. The other was focusing on cannons, armor, battlemechs, and so on, as this would allow me to capture more territory and hence improve my soul. I took pride in myself as an arms factory, and improving my soul was becoming a proxy for improving myself.
Luckily—despite the metaheuristic guided simulations not yet finding a competitive candidate for it—there were a few keywords that overlapped with both goals. I chose
mechs as my first manufacturing focus, as both my ancient battlemech designs and my internal worker mechbots should count (the battlemechs more so than the worker mechbots, which also used gearways).
Pride in myself as a mech factory was probably weighted higher than it should have been in that decision.
Well that was totally ridiculous, every last bit of it.
I gave it a whirl, swapping the production of a mechbot leg for a mechbot torso.
It exceeded my expectations. Even though it meant half completed gearbox assemblies being accelerated through the air to be split by a chopper in mid air just so... before ricocheting into a stuffer... after which they ended up as nearly completed gyroscopes instead.
This novel's true home is a different platform. Support the author by finding it there.
It made no sense. No matter how I tried to analyze what was happening I couldn't. My gears just knew what signals to send through the factory relays, it took barely any effort of computation on my part. I just willed it to happen and it did. I regretted not better testing the previous version of the upgrade now, it would have given me another data point.
More baffling than how I did it was what was actually happening in the factory. Gearboxes and gyroscopes were technically kind of similar, in the same way that say a wrench and a screwdriver were kind of similar, which was to say: not at all. Sure you held them in your gripper and they were made of hunks of metal. But one was a lever and the other an axel. A single chop and a little component stuffing shouldn't turn a force multiplier into an momentum regulator. And it definitely shouldn't be making progress manufacturing them as it did.
Baffled as I was, I regretted nothing.
So, a new focus every two ranks.
And manufacturing focuses had just become much more valuable.
I had been planning to choose
cannons next, leaning into my offensive capabilities. Even though guns always needed ammunition, I had suspected they were the safer bet. Now however the tricks with my new modular production lines tipped the scale the other way. I chose shells.
(The Manufacturing Focus keyword catalog allowed me to test objects, by thinking of them, against the keywords to see how well they matched. Apparently guns and bullets—which did not appear as keywords—were modern colloquial terms for hand held cannons and shells.)
This decision seemed like a big one. Behind—what I could describe only as—politician bullshit (and what was that doing in my soul) was a quality vs. quantity decision.
Push was very interesting at first, it might solve a near term problem of materials, factories were supposed to make a bunch of stuff, and potentially violating the rules of physics was tempting.
Pull however grew on me, it offered violations of the rules of information—which as an informational being first and a physical one only incidentally, was more enticing. And, I had always prided myself as a factory of quality machines, found no where else.
However what really decided it was the fact I was self improving. Every time I rebuilt myself it would be with better components. I chose Pull.
This did mean I would need to get serious about research sooner rather than later. I had always enjoyed research. Besides the fact those humans tended to be less stupid than the average, they had regularly presented me with challenges of manufacturing.
Now I got two? I wasn't complaining (about that part of it).
I had already decided to lean the other direction with my third focus, which was reinforced by choosing Pull just now:
manufacturing-machinery. As for the other I chose computational-components as it included relays—and also parts of mechbots, albeit with a small similarity bonus, probably my first overlap. And it was another self improving component, which with the Pull decision, had in general, become more favorable.
It was also because of some new Mastery Perks that had appeared.
These were somewhat interesting because of when they had unlocked. I was narrowing in on the soul's "direct knowledge" rules. Divergent appeared at Manufacturing Rank 1, before I made the choice. Focused Manufacturing Synergies appeared at Manufacturing Rank 2, before I had chosen the Focus (the upgrade it referenced had always been unlocked).
Focused Manufacturing Synergies was very interesting. Primarily because it provided multiple immediate benefits. I assumed a number of entries were missing from the list due to direct knowledge rules, however I would be perfectly satisfied with the existing list. I chose Focused Manufacturing Synergies as my last Mastery Perk for the moment.
It also displayed one of the most direct comparisons between the power of an—admittedly pretty weak—upgrade and that of a perk I had seen yet.
It was time to revisit my manufacturing priorities.
Stats:
The Factory Intelligence
-
-
-
- Caverns: 1,000 / 1,200m2
-
-
- Upgrade Points: 0 (+0.4/day)
- Perks: Specialization [Manufacturing], Focused Manufacturing Synergies
Project TITAN Factory
-
- Area: 4.2km2 (across 8 floors)
-
-
-
Manufacturing Rank Upgrades: Modular Production Lines II, Pull
- Rank Bonus: 12% throughput
- Focuses: mechs, shells, manufacturing-machinery, computational-components.
-