When the first dozen mining rovers came rolling towards the Lunar recycling and heavy fabricator building to unload their loot, Ash's main thread of awareness was working in a fully immersive virtual environment designing the first prototypes of five types of satellites his first satellite constellation would consist of.

Ash discovered that although he could just mentally update designs through an intuitive understanding of the CAD software, he achieved better results if he visualized the design in a VR setting.

To "see" the design, make a change and then "see" the result made his iterative design process much smoother. Maybe that was just because he was an amateur; he wasn't a real engineer; after all, Ash was just doing the best he could do to fake it. Ridiculous processing power and the assistant he fabricated tended to help smooth over the rough spots.

His assistant was an engineering specialist class 2 AI. Ash had fabbed her a mostly organic body, albeit with additional compute resources and transmitters to support wireless connection to a VR environment.

She did not have much in the way of personality but was excellent at pointing out where choices Ash made went against the standard and accepted engineering principles of the future and, to a lesser extent taking directions from him and creating an entire solution based on that, so long as it wasn't incredibly complicated.

For example, she was currently optimizing the designs for the Alpha, Bravo, Gamma and Delta satellite types to change their composition to maximize the use of materials available on the Lunar surface. While Ash was smashing his head against the much larger and more complicated Omega satellite.

The first three were simple projects, essentially just stuffing commercial off-the-shelf datanet, GPS and surveillance satellite guts into a stealthed satellite chassis. The Delta satellite was a little more complicated, but still, it was just a commercial fusion-powered power relay that used complex mirrors and high-energy lasers to provide power to the rest of the constellation.

The Omega satellite, though, was a complete invention of Ash's, albeit each individual component already existed, so it was more akin to putting legos together than an actual invention where a person started from scratch.

Ash wanted a satellite weapons system that could fire on Earth. At first, he went along the path of using a high powered laser, which would have been relatively simple to engineer as it was an incredibly mature technology.

But he came to realize something after doing some research on the subject. Something most sci-fi writers that utilize laser guns rarely tell you or don't know -- firing a high powered laser through the atmosphere would permanently blind everyone who saw it!

The coherent backscatter of light off air and water molecules in the atmosphere resembled shining a flashlight at a disco ball, and even just seeing the reflection of a laser powerful enough to do physical damage from orbit would cause permanent, irreversible blindness. Mass blindness, for kilometres around the impact site!

Even using invisible wavelengths of light near the visible spectrum like UV would not help that much; it just meant that the people who were struck blind had no idea why it happened!

While it was true that the main reason he wanted these Omega satellites was to have the option to cow hostile humans, along the lines of Tim the Enchanter from Monty Python, he did not want to strike everyone in a city blind!

Although he had to admit that it would be pretty impressive, but the message he wanted these to send was more along the lines of "Don't mess with me," rather than "Flee, flee for your lives! The great terror comes!"

Ash's first thoughts were to just change the part of the electromagnetic spectrum he was using, to back away from the more energetic visible light spectrum and UV/infrared range and use the range radar and radios use as that wouldn't blind people at all.

But weapons applications of MASER technology was somewhat limited in the same way that blowgun technology had stagnated in his past life once firearms were discovered. Or the way horse buggy technology came to a halt when cars were invented.

The closest thing he found was not in his design databases at all but a University paper suggesting that the large and powerful phased radar arrays mounted on capital ships, which was comprised of millions of tiny electronically-steerable antenna, could be constructed with tiny Vaslago superlenses built over each element of the array.

That would allow you to electronically steer the beam and focus the entire 100+ terawatt radar output down to a very tiny spot, smaller even than two times the wavelength! By Ash's definition, all that energy focused on such a small area would essentially be a weapon of mass destruction.

A great idea, Ash thought, and perfect for his needs! He didn't need anywhere near that level of output, either. Sadly, the facts were that Ash had enough difficulties incorporating fully completed engineering designs into a project. He was far, far away from the level of taking a research paper and spitting out a useful end product.

As for going the other way down the electromagnetic spectrum, to the more energetic wavelengths instead... well, that was just out of the question. X-ray lasers were enormous and complicated. Most research in that direction was along the lines of using the output of nuclear explosives to bomb-pump various lasing rods to generate an X-ray laser. And gamma-ray lasers were still in the realm of science fiction in the future; although there were signs that first-rate states have working models, they would be on the level of doomsday weapons.

Even if Ash could build one, which he couldn't, he didn't want to shoot a beam at someone on the Earth and then have it travel through the mantle and core and accidentally kill a city's worth of people on the other side of the planet! Or ionize the atmosphere by accident, or something else terrible.

What Ash had finally decided upon was a downsized linear accelerator. The technology to accelerate small masses to relativistic speeds was very mature and a standard weapons system on warships. They usually were spinal mounted cannons that accelerated a tiny bit of osmium or iridium up to 95% of lightspeed. In the common vernacular, they were called various things such as snowflake cannons or pea shooters. They were such a mature technology that most of the innovations recently were novel ways to cool them after firing to increase the rate of fire.

However, just building one of those weapons and pointing it at the Earth would be exactly the same as just bombarding the planet with nuclear bombs, albeit somewhat more ecologically friendly. So he might as well use the mass blindness ray at that point.

Instead of accelerating the tiny projectile up to .95C, Ash's version would at most get it up to .10C, and he wouldn't use such dense pellets, either. That was still a borderline weapon of mass destruction, so he downsized the projectile mass as well until the effective energy delivered on target was dialable to anywhere from 10 to 500kg equivalent of TNT.

That was what he wanted. A range of options from "Scary!" to Airstrike. If he placed these weapons platforms in geosynchronous orbit, it would likely only take between five and ten seconds from when Ash said go until the boom happened. Anywhere on the globe! Talk about Prompt Global Strike. America would have been SO jealous if they existed.

Still, it was a lot harder to miniaturize these devices than he thought. It wasn't likely the Omega satellites would be ready any time soon, but if there was one nice thing about being a machine, it was that you were patient and relentless. Failure had much less of a sting to him than he remembered when he was a human.

When a person had the capability to spin off a hundred parallel parts of their awareness to systematically try every single option in a problem, failure became just another datum. The more failures Ash experienced, the more excited he became because it was that much more likely the next attempt would be a success!

In Ash's new philosophy, what he was privately calling the Machina Dao, everything, the entire universe, was an iterative process. So long as Ash knew something was theoretically possible, then he would achieve it, eventually. That was certain.

Still, even relentless machines took breaks. Ash sat his work down and glanced over at Bingqing, his research assistant, "How's the status of the optimizations? I'd like to start a run of the surveillance constellation as soon as possible."

Bingqing replied, "Of the Alpha through Gamma, on average, we have achieved 62% composition of Lunar components, with the surveillance satellite a little higher. We have made heavy use of titanium and aluminium, but as you know, the stealth features of the satellite chassis are almost entirely carbon-based."

She shrugged her shoulders a little before continuing, "Additionally, since the intended orbit is so low for all of these satellites except for the Delta, we are unable to use any calcium-conductors in the microprocessor architecture -- the atmosphere at 200 kilometres above Earth may be thin, but it is still there, at least enough to compromise calcium, anyway.

I've created variants that use copper or silver conductors in a silicon semiconductor substrate, depending on which sources of metals we get access to. I think we can gain another 5 to 6% by replacing the traditional power cells with variants that use a silicon-based nanowire as an anode and manganese oxide as an electrode, both of which are in plentiful supply."

After tilting her head to one side and pursing her lips, she continued, "We could eliminate the use of graphene-based conductors and save carbon that way, although the power-cells would be inferior."

She clucked her tongue and finished with, "After that, we are well into the realm of diminishing returns, I think."

Bingqing chose her own name when she was first spun up after fabrication was complete, which was a little surprising. Ash felt that the class 2 AIs in his database were on the verge of being considered class 3's in terms of the complexity of their neural networks.

Although it might be correct to say she did not have much of a personality, another way to say it was her lack of personality WAS her personality. Most specialist AIs, even 3's and especially engineering specialities, were on the autistic spectrum anyway. It was what happened when you hyper-focused a person's interest onto one thing.

The textbook answer of the main difference between 2s and 3s was, beyond the fact that 3's were developed using the patterns of human minds, was that class 2 AIs didn't have much conception of self. As a result, they could be left unattended indefinitely without any real negative consequences, whereas a class 3 AI would suffer severe psychological problems in the form of loneliness over time.

For example, the ship's AI seemed lonely when Ash came aboard, but the textbooks would say it only expressed that loneliness in retrospect. The traditional opinion was that class 2 AI's only expressed emotion when in the presence of emotive entities such as humans or higher-tiered AIs. Ash wasn't sure he bought this theory, as it seemed an excuse to treat class 2 AIs poorly. Didn't his loyal doggo Shapur get lonely in his old life when Ash went to work?

That was why one of the six AI's he has fabricated so far had her main purpose being to keep the ship AI company. Technically she was in charge of logistics and putatively in command of Mistress of Space when Ash wasn't on-board, but the reality was he built the ship's AI a friend.

Ash nodded, "That sounds good; we can accept the slightly inferior power cells. The constellation flight plan is set up so that they're almost entirely within LOS of one of the Delta sats in a higher orbit, anyway. Get to a stopping point after you incorporate those new power cells, and we'll run off a few prototypes for physical testing. Let's not chase sub percentage point optimizations at this phase. By the time the service life of this first generation is completed, we won't have the same supply constraints."

Bingqing indicated her understanding, and Ash disconnected from the VR space to find himself looking up from the bed in the Captain's quarters aboard his ship, the Mistress of Space IV. Although the truth was that he was watching with his body's senses the entire time, Ash has formed an opinion that his "main" thread of awareness was the most important one.

Although each of his parallel threads was indisputably him, he had noticed that his consciousness marks one as more important than the rest. A first among equals, or to use a computing term a master node of a cluster. The rest operate at a level below his consciousness but still above his subconsciousness -- he can immediately know what was happening, and if one of his parallel threads comes across something specifically important, it was usually elevated to the main thread and his consciousness shifts to that perspective.

Ash thought about it a while and finally came to the conclusion that that was probably an intentional design choice; otherwise, Ash would be some kind of gestalt hive-mind entity. Humans of the future had a long tragic history of dealing with a particular species of aliens that had those characteristics, and the idea of creating a similar AI probably scared the crap out of them.

Ash wondered if it was considered a Xenocide if the entire alien species was a single consciousness, or was it just considered one murder? He had a medium to long-term need to know the answer to that question.

Although, in truth, it was academic because even if it was still a great sin, he would nip them in the bud before they had a chance to xenocide three other alien races and kill trillions of humans. But those were long-term plans. This species did not become a threat until they were exposed to and reverse engineered FTL many, many years from "now."

Thinking more about AI rights, he was also considering building the ship's AI an actual body, too. However, the boyish Ship was an extremely complicated AI, even if he was technically only a class 2. Ship featured a highly parallel architecture, although nowhere near the level Ash himself had. The chassis would need incredible networking capabilities for him to continue working as he had been.

It would be essentially impossible with his current tech base to include that in an organic body. And only a handful of the synthetic body options would work too.

His tentative plan was to use one of the higher-end synthetic body options but combine it with the serious networking, and auxiliary computing resources Ship would need. This would be a bespoke solution and would be a small project to pull together, which was why he hadn't begun it yet. Another option would be to build a body that Ship could operate like a drone, remotely, but it didn't entirely sit well with him.

Although Ash's plans were to make heavy use of drone bodies, the longer Ash lived as an AI, the less he liked the idea of being genuinely disembodied like Ship was. Ash would have to talk to Ship about it, though. But, on the other hand, perhaps Ship thought the Mistress of Space WAS his body, in which case having a remotely piloted drone body to interact with people might be acceptable.

If that was the case, wouldn't Ash have to change the ship's name to the Master of Space?