chapter 61
Building the new personality cores was quite exciting. There wasn’t as much need for most of the AIs to be too complex. The normal-sized personality cores fit into my hand and were about the size of a baseball. These new ones were a bit bigger than marbles. This also meant they had a smaller amount of stored power inside, but that was fine, as they also used less.
We would also need a lot of high-capacity batteries. Fortunately, they were quite cheap and easily purchased. We were going to need quite a lot of drones, as we needed a pretty big workforce if I wanted to finish building this ship in a reasonable amount of time. Most of these drones would also double as maintenance drones after I finished the build.
For the next nearly two days, that was all I did. Build new personality cores for my future workforce. Normally, I could have finished this job a lot sooner, but I was really diving into how these new types of AI personality cores worked, as I still had quite a lot of already built AIs that I needed to transfer into these new cores. They were simply so much better than what I had available before.
An interesting thing I found was that I could have nested AIs. At first, I aimed to have 100 new ones, as that would take me closer to my theoretical limit of how many I could be bonded with.
Of course, that limit might be higher, as my affinity might affect it, but I think I found a way to have a lot more than I expected.
The nested AIs didn’t work in the way I expected them to. What I expected was to have two fully independent AIs, with one acting as the superior of the other. This was something I wanted to have for a more robust command structure. But this discovery changed a lot.
I could have completely separate drones, yet have them act as almost an individual. There would be a superior, so to say, and the nested AIs would not be individuals but more like parts of its body.
They would work like your hands, having somewhat of a mind of their own but still following what you say they need to do. This way, you could have a group of 12 drones doing completely different things, yet still only be one bonded AI.
There were limitations. If every one of those drones was trying to do too many different things too far apart, the connection would strain. But of course, there were ways of mitigating it. The simplest being making groups of drones that did the same thing be governed by this bonded AI group.
Theoretically, the more complex the governing AI was, the more AI-controlled parts you could add to it. This opened up so many more pathways I could build things around.
I was even thinking that maybe some of the more complex parts of the Mark II could work more efficiently—and better—if they had their own AI running them, with the governing AI being FH herself.
There was a lot to think about, but all the processors had arrived, so it was time to start building out the future server and computing unit of the Mark II.
The current housing of it was quite a simple box, but when we’d install it on the Mark II, it would have quite the reinforced housing. It took me a week to finish setting everything up, the hardest part being managing the temperature and power input.
Right now, there were huge cables coming from the ship to this server box. It looked quite janky, but it had nearly 100 times our previous processing power. Now we could run more complex VR environments, and of course, we had enough processing power to properly run structural simulations for the Mark II.
Time started to go by faster and faster. Designing the drones and droids I would be needing for the build was fun—in fact, everything about this build was fun. I quite enjoyed building the Mark I, but there was always the feeling of doom as I did everything. Now I could just purely enjoy designing and building.
In only three weeks, I had myself a workforce of nearly 400 drones and droids, a lot of them the nested variety, always working in groups that stuck close to each other.
The Mark I was stripped down: all the armor removed, the engines gone, and almost everything disassembled, with only some crucial pieces still working, like the power core.
The remaining structure was now up against one side of the wall so we would have free space to start building the structure of the Mark II.
The simulations were finished, and it wasn’t as bad as I expected it to be. We still needed to use quite expensive structural materials, with the current cost of the framework coming to 137 million.
If you encounter this tale on Amazon, note that it's taken without the author's consent. Report it.
That was basically the cost of the smallest working FTL-capable ship you could purchase, but this was just the basic structure for the Mark II. Although it was incredibly important as it was responsible for holding everything together.
It would take a bit of time for all the pieces to be made and brought here, so it was time to finish upgrading my AI.
During all of that, I also worked on understanding the personality cores. What really helped with that was getting my 6th Sense to the Apprentice rank, but that wasn’t the only skill I managed to upgrade.
Mind Guard and Telekinesis have also upgraded. Yet the most important one was 6th Sense, as now it worked even better, helping me understand the personality cores even deeper.
The result was that it was time to start upgrading the simplest AIs I had and transfer them into the personality cores. The process was quite nerve-racking. From the old computer core dozens of wires ran to a half-finished personality core. All of this looked kinda janky.
“Are you sure you want to do this? It could mess up their personalities,” Lola said.
I was also quite worried, but I needed to learn to do this so I could give Lola a better way of interacting with the world than having to move around in a huge mech.
The transfer itself was quite fast. When I finished assembling the personality core and awoke the AI inside, the relief I felt when everything seemed to come back green—and when the personality test finished with exactly the same results as before—felt like a huge weight was lifted from my chest.
Of course, we needed longer testing, but slowly I started to transfer more and more simple AIs I had from before into the new personality cores.
I seriously believe that I wouldn’t have been able to accomplish this without the 6th Sense skill. It was absolutely invaluable. This allowed me to see what skills could help me accomplish, which made me even more excited to find more skills and see what they could help me do.
The first structural pieces started to arrive. FH had come up with a way to more securely bond them than the stitching tool would allow.
Basically, we would use specially made nano machines, a lot of heat, and some pressure to weld these pieces together. This bonding would not only happen at the connection point of the two structural pieces, but further into each piece.
The nano machines would burrow their way quite deeply into both pieces, quite similarly to how the stitcher would work. Since we could make the nano machines mostly composed of the material of the two pieces being combined, the connection would be almost perfect.
This would make the base structure of the Mark II more resilient. We actually didn’t need this type of connection for the structure to work, but why not build the best we can? It was amazing to watch so many drones work together to slowly start building out the Mark II framework.
As they continued to do that, I focused on understanding AI personality cores so I could start transferring more complex AIs. I wasn’t going to transfer any of the more complex AIs unless I was 100% sure of the outcome. For now, I still had plenty of learning to do, but everything seemed to be coming together nicely.
As I had Bob in front of me, I started to think about transferring him over. Once again, as always, my 6th sense triggered. A week before, when I tried this, I got a sense of wrongness, like doing so was a bad idea. This time, the only sense I got was a slight adjustment, just related to some electrical connections I was planning on making for the transfer to work.
Over the course of this time, I had learned how my sixth sense worked and what the feedback I got meant. When I got an overall feeling of wrongness, it meant that I didn’t know enough and that doing so would most likely end in failure. When I got a sense of how to do things more efficiently or just cleaner, it meant that what I was about to try would work out.
During this time, I had learned so much about how AI actually worked, and more specifically, how the personality core worked and its differences from more traditional Earth-based AI housing.
Whoever came up with this personality core was a true genius, and even during all this time with it, I had not found a way to make it better. Although I did find ways of changing it, most significantly by making it smaller for simpler AIs.
Bob was also quite the special case, as even I didn’t fully understand what made him tick. I was confident in what I was about to do, so I proceeded.
Bob also seemed quite excited because he understood that he was going to get more complex after this, and even though he shouldn't have understood it, he somehow knew what that meant and wanted it.
Slowly shutting him down felt wrong. Not in a sixth-sense kind of way, just in the way that shutting down an individual feels wrong, even though I knew I could just bring him back by powering him up.
What followed was almost three hours of wiring different connections to a personality core that was twice the size of a normal one.
Unfortunately, there was no way of fitting him into a normal sized core, as there was so much unnecessary code, but I couldn’t delete any of it. Tampering with even a bit of his personality would result in total loss of his personality matrix.
The transfer itself was, once again, fast. As I finished assembling the personality core, everything felt right.
When I powered him back up, everything seemed to work. But when it was time for the personality test, things started to feel off. The answers were correct, but he took way too long on almost every question. Did I mess up his processing capability somehow?
“Bob, is everything alright? You seem to be taking a lot longer with the personality test.”
“Well, these questions are badly structured. They need to be altered, and not being able to do so is annoying me.”
That stunned me for a moment. He wasn’t supposed to be able to talk like this, at least not just yet.
“How are you able to speak? Do you feel different?”
“Well, that’s a stupid question. I was inefficient before, so I fixed myself as I started to wake up and saw the absolute mess my processing logic was.”
That’s right, Bob didn’t have any internal limiters for changing his own code.
“You need to be really careful about that. Changing just a bit might break you.”
“Don’t you think I know that? But there is organization in my disorderly code, and I will never mess with something that works. I just needed to clean some pathways. Now I will finish this test, trying to overlook the badly structured questions. Like this 17A—the question itself has three different ways of understanding it. Fortunately, the answers themselves help make sense of the question.”
I was completely dumbstruck. For some reason, he felt like Bob, even though he was a lot more complex now. Did I make a mistake in putting him into a personality core?
“Now that that’s done, Lola, my dear sister, I have some ideas on my future body to make myself more efficient in the organization of the Mark II.”
“Of course, Bob, I’ll be happy to,” I heard Lola answer in a genuinely excited tone.
“As for you, Captain. The way you have designed the current cargo handling of the Mark II is... lacking. You have wasted so much space, and what is this single rail lift? It's way too limiting and inefficient. I will start designing some proper rails for cargo management, and we will then see what changes we need to make.”
Oh no. Of course he was going to be like that, especially about cargo.
Even when he was a simple AI, he had a drive I didn’t understand, about how to properly manage cargo. Of course that would get amplified to no end now that he was more complex, as the base of him was still the same.
So it was a success. Although I had a feeling that now I had quite an overbearing quartermaster.
No matter how annoying Bob might be in the future, this brought a true smile to my face as I listened in on him discussing with Lola about his new body.