---Richard---

Richard was making good progress.

He was far enough into his secret project he was relatively sure he actually complete it.

The project was simple. He was going to make computers and servers. The internet. And finally Skynet.

This project was based on what he had learned from the last area. The final fake zone with its equipment that hide a secret area at the hardware level.

This was Richard's one shot. His one and only shot to fix a problem before it appeared. It was honestly pretty simple of a goal. Richard refused to let this fun new world devolve into a second earth. The easiest way to do that was to set up a system that killed corruption and secret abuse before it even started.

If he set up a proper internet before anyone else, made it integral to everything before people found out what he had done…hid what he had done as well as he could in the hopes no one could discover.

Well then just maybe the idea would stick. His only shot was to make this system first and make it better than others. If someone else made one first his would never pick up steam. If his sucked, it would push others to make a better one.

So computers. How would you go about setting up a computer from scratch? What would you consider the most important point to focus on? How would you even start?

There is frankly a massive amount of interconnected parts to worry about. The only way you can even begin is to compartmentalize and focus on as simple an area as you can. The information you prioritize and the base structure you head towards cascades and changes everything else. Even before you start thinking of earth architecture with cpu’s and motherboards and ram, you have to think about the base building blocks. Transistor equivalents. Clock equivalents.

To build off of what was done before Richard looked into phase crystals. They were the building blocks for personal AI’s. The specialized components for information storage and calculation as he had learned in school.

They had a documented well known “recipe” to create them. There had been patents and monopolies in the previous zone…but those didn’t matter in the real world Richard found himself in.

The ‘Recipe” was as follows.

To make a phase crystal you first needed to bind single atoms or molecules with high energy light. The element you decided to entangle with light completely changed every calculation past that point most requiring entirely different machines to even test.

The bare minimum for such a machine was to create a perfect vacuum chamber filled with aerosolized aether primed atomic dust – AAPAD for short. Low wavelength aether primed light could then be shot into the chamber exciting the elements within into a vibrating high energy state HEAD for short. The chamber itself needed a mirror equivalent to keep the light entangled atoms bouncing and an anti-adhesion equivalent to prevent the atoms from fusing to the case while in this stage. If the base molecules used were water for example, a reflective and superhydrophobic coating was usually enough – especially one you could boost with aether for a short period of time.

Water was sadly one of the worst options for what Richard wanted as its specific shape lent to biased phase crystals no matter what else you did. Chirality broke most phase pattern even worse than something like water and while certain chiral molecules were strangely useful proving exceptions…most just broke everything and were a dead end.

From what Richard could tell the chiral elements that didn’t break things shifted into achiral shapes when excited with light entanglement anyways, so they didn’t fully count.

Using the same logic base molecules with 6 bonds like ditungsten gave some of the nicest and most stable 3 phase shapes…and stuff like methane with its 4 bonds evenly distributed were incredibly stable as well in all the ways that mattered. Lots of carbon-based molecules worked great but the more complicated they were the harder they were to work with or synthesize.

This sort of logic mattered less and less the more aether you used. Water stopped biasing at ‘rank 2’ levels of aether for example but was still not effective for other reasons.

So back to containment.

Something like cesium or calcium chloride or even something like benzene required entirely different methods of isolation to water and the method used could not react with the components being created. Kinetic barriers worked for half the elements but broke the other half in specific ways. Some elements were perfect as far as his AI could hypothesize but were impossible to work with.

Pesky real world getting in the way of science.

The next step was equally as simple in concept but had an even greater number of factors you needed to keep track of.

As soon as your central excited ‘core’ was created, you needed to completely enclose that entangled point in an insulating ‘crystal’. One that perfectly insulated the central core and preferably had certain measurable properties to interact with it. The easiest method of creating this shell was to blast them all with a source of decaying entangled light atoms – the unstable variants ‘breaking’ and depositing their elements on the outside of the inner core. Basically, you could coat the floating cores with the equivalent of an atomic spray paint.

That was the easiest but not the most effective. It was fast but usually resulted in uneven coatings and lumpy shells created phase crystals that biased towards certain phases.

That bias happened a lot.

A more effective method was to create a sort of liquid crystal compound, and then fill your vacuum chamber with that, allowing the phase crystals to grow slowly. This method made much more evenly coated phase crystals and had fewer side effects than the spray paint method which usually had bad interactions between the spray paint light and the entangled light…but if you went with this method you needed to figure out how to both create a viable crystal bath and a safe way of injecting it into your chamber. Simply opening up a hole into a vacuum would suck it all in with enough force to break the fragile cores so the liquid coating needed to be kept at an incredibly low pressure and slowly released in over several minutes.

All of this ignored the actual effects that these cores and coatings had.

Certain elements resonated at higher or lower frequencies. Certain elements decayed or released various radioactive byproducts. More important than those properties was its overall stability. Certain elements were mostly stable at low aether levels while other elements needed continuous flows of high tier aether to stay together. Some of the core-shell interactions canceled each other out making expensive rocks while others synergized in unwanted directions making crystals that exploded or even imploded or disillusion popped themselves when prodded or measured.

The elements that functioned at Richards max aether flow were many times “faster” and “accurate” as computer components then the low flow elements…and many of them had the most information on them.

The components used in personal AI’s were all ones that needed a steady flow of short ranged aether just to function after all and they had as a rule truly minuscule footprints. If Richard wanted to create a server that could run without someone standing there he needed to focus on low-flow stable elements and ignore most of the existing information completely.

To make this whole process even more fun even if Richard put his all into it, he just didn’t have the resources or power of a personal gravity reactor to even attempt most of the stronger high yield builds. His attempts all created phase crystals with a high degree of failure – the bottom of his chambers filling with a glittering dust that had maybe 10-30% proper PC’s and 70-90% duds. Trying to test or clean up his failures was a whole ‘nother mess. It was better to ignore it and create components that self mapped ‘good’ crystals while ignoring the bad ones.

Now. All of this was a lot of work and completely ignored completely aether free components. Transistors, Diodes. Resistors. If they were good enough for earth they should work here too right?

The problem with those components was a combination of scale and efficiency. Even the “worst” and “slowest” phase crystals were usually on par or better than the earth variants. Even if they performed ever so slightly worse, phase crystals had 3 channels of information to a binary’s 2 and his AI has so many algorithms and architecture blueprints that relied on this 3 bit base.

The other option was to invent something completely new. Randomly come up with an efficient transistor equivalent from scratch. It was tempting. Oh so tempting.

How would you even start? Randomly bang your head against the wall until it works?

Richard did actually test some options along that path. He had an entropy affinity even if he’d scrubbed it from his skill slots. He could alter aether parameters back to that base to try probability manipulation.

And the ‘thousand monkey’ style of random chance was helpful in a few case. Richard ended up giving it up after realizing how long it would actually take however. He tempered his expectations for easy solutions being handed to him and came back to studying phase crystals and noting down ones that might match his use case. There were billions of potential molecules after all and even if there was hundreds of strange exceptions to soft rules you thought you had discovered, you could map out patterns and trends. His AI was great at graphing trends and guessing potential materials even if it was wrong some of the time.

…

Richard was constantly checking and mapping out how different phase crystal bases might help but focusing on the building blocks for too long would never let him complete this in time.

So building upwards what are the most important parts of a computer? Abstracting a bit there was the main clock obviously. A method of ‘ticking’ commands to the hardware and reading those commands. You wanted your clock to run as consistently and reliably as possible. For this Richard discovered a specialized clock crystal that was built a bit like an elongated phase crystal. Just saying everything was a PC discounted how different this element was – it was created by light infusing a long polymer and crystalizing it into as straight a line as he could.

Enjoying this book? Seek out the original to ensure the author gets credit.

Basically a wire with a chunk of complex energy bouncing up and down it. The longer the wire the slower the clock frequency. The straighter it was the more reliable. No matter what the shape was the phase would be able to correct itself but if there were curves in its line, the phase would wobble and create ticks in sinusoidal patterns instead of even ones which wasn’t ideal. Useful for certain types of hardware frequency analysis but hard coding a microscopic Fourier transform component felt like getting distracted.

These clock wires were many times more fragile than regular phase crystals but vital to build off of. Richard noted them down and moved on.

The second most important component was a method of reading and writing values.

Phase crystals only worked as a storage medium if you could read the phase that was bouncing around in them. At incredibly high levels of aether you could ignore more and more requirements – something like rank 4 aether would let a pile of phase crystals with nothing else in them start functioning like a computer.

One method required nano levels of architecture. Glue a measurement device – synced to the elements in the core but not blocked by the shell – on 3 sides of the phase crystal testing to see if it was currently in that phase.

You actually only needed 2 measurements – if you didn’t care as much about accuracy you could assume a lower measurement was a ‘not’ and ‘not a or b’ equalled ‘c’. 3 measurements let you compare which had the strongest measurement however and worked essentially 100% of the time without false positives like the cheapo version.

If you glued 3 ‘setters’ on the opposite side of the phase crystal you could force the phase into one of the 3 directions. How that worked depended on the type of crystal but usually involved some sort of inductive pulse to pull the phase towards it like a magnet.

That 3 getter, 3 setter per crystal setup was the most reliable but understandably hard to engineer at microscopic levels. You could head back to the spray paint method blasting up down left right front back as fast as you could with different materials…or figure out how to make nanobots that could build each cell.

Richard ended up outsourcing a nanobot solution to a friend of Muhammad. Each person who helped was a loss of control and potential leak but it really was the only option that worked for him and tied into the wiring as well.

Sometimes Richard thought he had been given the short end of the stick. Other people had more egregious cheats.

…

There was another method of measurement. One that didn’t require microscopic positioning.

It only worked with a very specific type of phase crystal. If the base was something like SCuAgAu – something only possible to create with high levels of aether – then you could remote set the phase by using copper gold and silver based ‘setters’. They would set all phase crystals in the same area to the direction of their relevant atoms and by using similar remote ‘sensors’ you could collect the average phase of the crystals in the same area.

This was so much harder to design around but suddenly became valid when Richard stopped caring about size.

Why had he been focusing so much on making a teeny tiny computer? What if he placed the components across several meters of space?

If you combined this mass set concept with the individual set, you suddenly unlocked several algorithms for certain kinds of problems. If you created an isolated device that took advantage of this…well then you’ve just created a chip like a gpu or network card.

…

Everything began to feel more and more esoteric. Richard made a complete miniature computer and then shifted it into pure energy attaching the computer to a second computer. The result calculated his test suite of problems in less than a millisecond and then imploded vanishing from sight as it was pulled into a higher realm of energy.

…

Richard designed a custom programing language. It had branches, fuzzy branches. Recursion, fuzzy recursion and a secret function that printed an ascii dick on any display imaginable.

…

Richard used his own blood to stabilize some of the unstable phase crystals starting the process from scratch once again. At some point his science had started feeling like mad science but Richard was confident the feeling would pass when he understood more.

…

Richard's second attempt at an energy attachment was done using stats. He permanently sacrificed one body speed – accepted the system confirmation to allow him to break it off – and then attached a simple computer to the one super heavy atom of X. This time it didn’t disappear into nothing after a few seconds.

He sent information to it. Received a response.

Waited a bit and repeated the experiment.

Computer is still here and accessible!

Now if only he could figure out where that atom went – it was somewhere on this Petrie dish…

…

There was an important item in all these experiments that was hard to ignore.

The mcgruffin. The alien cheat.

The nest egg.

It wasn’t some magical tool. Richard couldn’t press a button and have it print out a computer for him…

But it was still more powerful than anything else and could not be ignored. If he reset the nest and fed ‘perfect’ phase crystals cells into it while the nest was growing…well then in certain cases it would start appearing in the nest itself. Monsters with phase crystals in their brains and nerves. Plants that used simple logic to react or grew stems full of inactive phase crystals for no apparent reason. Monsters that communicated using remote activated phase crystals and formed hive minds with cloud consciousnesses.

Results got even weirder when you varied the inputs. It was impossible to feed ‘just’ phase crystals into a nest after all. Either it included air and stone and grass…or if you removed that it contained a vacuum. It always needed aether to grow and even a default slot had microscopic variants in the power frequency quality and flavor. You had to specifically create super filters just to get a 100% ‘pure’ aether and that meant most of Richard’s attempts had something like 0.1% of that entropic ‘flavor’ still in it.

And even when you controlled everything, something slipped through and varied your results. The most nightmare-inducing nests were the most promising. Richard's current iteration had some weird confluence between: his skill liquefied stone, embedded phase crystal cells (including various microscopic circuits), specialized chips he had made for dozens of different purposes…and germs.

Germs that had appeared from his hands or the air or something and had chosen this iteration to affect the nest.

This nest had two nightmare inducing horrors that Richard wanted to harness. One. A pervasive technological virus that threatened to take over every AI then entered the nest – potentially hosting itself on your AI and traveling outside of the nest afterwards as a sleeping trojan horse.

Terrifying especially when you realized Richard had offloaded control over his stats to his AI. All the super virus had to do is weaken his defense at the worse possible time and something could break him.

The only way to be mostly sure technology was safe was to completely power it down.

Remove aether and let all the RAM equivalents fail. Boot it back up and scrub the ROM for anything malicious. Rinse and repeat as the virus was adaptive and stubborn.

It was a calamity sure to cause a witch hunt should anyone else find out about it. It was also the sort of staying power Richard respected. He wanted his anti corruption Skynet equivalent to have the same kind of staying power. No one could ever scrub it from the system.

It was kind of funny. Fight one kind of corruption with another. Fire with fire.

The second horror was even more useful.

The mainframe for the nest – the ‘mother’ that controlled all its drone-like monsters and was responsible for attacking technology spread through stone.

That’s right. Slowly but surely the computer grew. It grew in stone – liquifying microscopic bit by bit and ever so slowly shifting the atoms around into more of it.

Like some grey goo. Like a von Neumann probe.

Burn the ‘source’ and it would pop up a day later somewhere else almost completely healed. Useful if Richard could recreate it. Self healing. Able to split into two and spread to multiple cities.

The final horror was the fact this self replicating technology corrupting calamity did not necessarily need to stay in the nest.

The ‘source’ and her stone burrowing drones kept to the nest because of some vague instincts but they stayed active outside of its borders. The longer Richard left it alive, the higher a chance of permanently effecting the world this nest had.

And yet this was the best chance he had of completing his goal. The tension made him feel alive and he kept giving himself more rules to follow. If any of his rules were broken he would shut it down.

Today was a harvest day.

Richard melted off huge chunks of stone computer – completely AI blind and harvesting the growing mainframe feeling somewhat like a beekeeper removing chunks of a hive.

He could be stung at any point. He probably wouldn’t be.

The mother learned and adapted and had long since stopped sending her minions to attack Richard. She knew it was useless and preferred to try and corrupt him. Either that or she was secretly growing some counter.

Probably that.

Walking to the perimeter fence he had set up, Richard stepped into his personal decontamination building and shook out the microscopic worms the mother had slipped into his stone boots.

This whole perimeter made him feel safer. He knew if the nest got serious it would just dig under his wall but he had set up enough sensors he was pretty sure he could tell when that would happen.

Placing the hive tablet in a conductive bath Richard reached out with a connection hand and blasted the bath with as much aether as he could channel for a full minute.

Overkill was the only level of safety he was confident with.

As fun of a MacGuffin as the nest was, Richard would never use the materials for it in his final product. He would never be able to trust the hardware or software to be free of the mother’s clutches. Wouldn’t it be convienet if he could just harvest this nest and put his own program in it? Just let the nest do the hard part of making everything.

No, all he could do was study and learn from it. Attempt to replicate it himself only after understanding all there was to it.

Richard took his block and moved towards an inspection room. He would study the mother’s wiped hardware just outside of the nest and then melt it to scrap.

Tapping into an isolated AI and slowly powering up his research room Richard had made just for this facility Richard was immediately attacked by a bolt of high energy from the corner.

It burned halfway through his face before the turret ran out of supplied energy. Richard swore again and again before pulling out some healing cream and slathering his wound.

How had that tech witch corrupted the facility already? He had taken so many steps to try and protect it. Richard had immediately cut off power immediately and was still hurt this badly from his own defenses.

Should he burn the facility to the ground and try again?

Stop poking something he couldn’t control and try something safer?

Richard placed the chunk of wiped nest in a new bath. He then left the room and building riding a low tech buggy down a path to a new building nearly a kilometer away. Entering the room he checked the security logs – paranoid as he could be – and then melted a hole into his perfectly sealed stone box before placing his regular AI on.

He had lost so much information from his volatile storage his constant companion becoming a backdoor into him. The AI connecting to his dedicated slot and booting up filled Richard with happiness.

Really. He should quit. Wipe the nest burn it to ashes and try something safer. Richard didn’t think he could keep doing this any longer.

A message from James!

Richard smiled slightly still full of stress but happy at the message. Opening it up he set the message to start playing with an audio interface then flopped down on a soft chair rubbing his eyes as he listened.

Slowly his mood soured even further. What stupid ideas were James’s current companions entertaining!?

Richard blasted off his reactions unfiltered then sighed as he started receiving responses. Of course James didn’t think it was that bad. No one seemed to mind kids as much as Richard did.

Richard kind of knew his stress was leaking into his interactions but fuck! All this work he was doing to try and make the world a better place and they would just leave it to a group of grimlins?

Richard continued messaging James and caught up with what they were doing.

They…were creating a calamity.

They were just doing it. No safety steps. Just balling it.

Wow…I’m being a bit of a pussy aren’t I?

Richard stared down at his hands.

His friend was having the time of his life making a world ending threat and he was considering melting away his own?

He had to match that energy.

Richard stood up and jabbed his fingers into his head allowing his AI to modulate the defense around the area. Slowly Richard ripped his AI out then carefully placed it in the box aligning the stone in a specific way he was sure was hard to emulate. He went over to the corner and grabbed his second facility AI wrapping it in a sling and heading out once again.

He was going to crack the secrets from this calamity yet see that he didn’t.

…

Richard returned to the lab and installed his newest AI pulling the old one out and smashing it to the ground.

This time when he powered it up, the facility didn’t start attacking him.

Using the machine to start activating a series of sensors Richard projected an image of the hive up onto the board behind him.

Low tech. We are back to projectors.

If Richard had his personal AI, he could display it in his vision as an overlay.

This would be worth it.

Pulling out the first of a long series of sensors Richard got to work.