Artifact Page 21
“Projects,” I said, coming back to focus.
The vault’s time–lock – which had become a symbol of security and safety since Patrick’s death – suddenly popped open. Sarah and I both looked toward the sound.
The eight time–sensitive locks finally reached their window. If there was anything on the other side that had the combination to open the vault door, this was the time to do it.
The thudding finally stopped altogether, and then the combination dial slowly, almost imperceptibly started turning.
5.
[SITE.const], [excavation.] [LAUNCH], [Neurl.limbic.archi], [MICROelectromech.systems], [HIL.lex], [Onsite.artifact]
Right after the time–lock snapped open, both of the combination dials started turning slowly in opposite directions. My first guess was that the thing on the other side was spinning the dials, but at second glance it appeared that they were moving in a more mechanical fashion, rather than being manipulated by something with a specific goal in mind.
“It can’t get in here,” Sarah asked quietly. “Can it?”
It must have been some probability function, which decreased the likelihood that someone, or something, could stumble upon an already impossibly diverse combination, by spinning the dials and making them more random. It really depended on how many digits the combination was. I figured that a twenty nine digit combination lock had a one in twenty seven thousand chance of someone randomly finding the right combination. This vault’s combination, however, looked like it could have been a one hundred digit lock, or even a thousand digits. The odds of randomly opening it were ten to the nth removed decimal of likelihood. I assumed that the numbers being in constant motion just decreased the odds even further.
“I don’t think so.” I said finally.
“That man said it was going to eat your brain…?”
“Patrick? Yeah, he did.”
“So that it could know what you know?”
“Yes…”
“Well,” she said. “What about him? Did the monster learn what he knew after it ate him? What if he knew how to get in here?”
That was an interesting thought. I watched the locks for a moment until I was confident that the spinning was some sort of automatic process. “Keep watching it, will you?”
She nodded.
I turned back to the wall–screen. What was going on? I could guess that the Research Center had various projects that they were working on, many of which probably involved various government contracts they would have wanted to keep under wraps – that’s reasonable. But it seemed that the CEM was bankrolling some serious stuff. I studied the file names. Onsite artifact…
That’s what I was looking for. “Onsite artifact.”
WEEKLY ACTIVITY REPORTS
“Here we go,” I said. “Weekly activity reports.”
The screen zoomed out as thousands of entries filled the wall, floor to ceiling. The first entry was dated May 15th, 2034. I walked to the right, scanning the various dates until I found the last report at the very bottom of the wall, which was written on August 19th, 2042. There were thousands of reports. Way too many for me to try sifting through each one, so I decided to start with the first entry and work my way randomly through the list, until I finally reached the last.
Week Ending 5/15/34
Classified Restricted Ivory Tower Redshift
Lancelot Kattar, Center for Energetic Materials
This first week wasn’t without hiccups. The accommodations were perfect, no less, but aside from braving the crowds of homeless, we had a brief but unsettling correspondence with a research group from CERN, who cautioned that they suddenly found evidence that implicated what screwing up would mean at this point – if we are even one logic gate off, we would most certainly short circuit and ruin the electromagnetic field of our wafer, which would ultimately set us back to square one – at this point, after so many man–hours, I can’t really imagine what such a failure would feel like. They’re right after all, but Jesus, I don’t think they realize how important everybody’s mindset is at this point. If we started focusing on what failing would mean, it could prove to be such a distraction that we’ll bring about the very failure that we’re worried about. I understand that we were basically told that the powers that be have amassed near unlimited resources for this endeavor, but I can’t just shut off an upbringing of despising wastefulness and praising frugality. If only the rest of the world knew. If only they could rally behind our cause and bolster our confidence so that we could do this right the first time. I would hate to waste another score trillion dollars in an equally risky project.
Another complaint I have is the complete lack of developmental perimeters. The project managers come to me and ask me to set aside what technological hurdles we face from now until then, and tell me to simply focus on the end result. What do I want to happen when this is all done? My answer is always the same – if I knew what sort of budget I had to work with, what sort of equipment we had, what sort of equipment we could get our hands on, then I could plot out a definite plan of attack. I could know exactly what the end result was going to look like, if I knew what tools I had to get me there. They say the same thing, that money isn’t an issue, which sounds accommodating but is actually quite stifling – if budget isn’t an issue, what would I conceivably need to get the job done? If we could conceivably get our hands on any available resource or equipment that I need, what then? If the technology doesn’t exist yet, what materials would I need to make it? I find that I’m suddenly paralyzed by the enormity of choosing from an unlimited list of possibilities. Screwing around in campus labs was almost easier, because I knew what my limitations were – I knew what things I had to work with, and what things I didn’t. Suddenly they tell me that I have the entire universe at my fingertips, and that all I need to do is ask for it. Having unlimited options happens to be very intimidating.
Week Ending 10/23/34
Classified Restricted Ivory Tower Redshift
Lancelot Kattar, Center for Energetic Materials
During the initial geometric etching, we were able to program a set of algorithms onto the first sheet of rudimentary microprocessors with a subtractive method, similar to how sculptors shave sections out of a block of marble to create a shape, rather than constructing shapes by building an object from the bottom up, with smaller components. In other words, we started with a large block of something, and carved away the unnecessary parts. It’s a process of construction by destruction. What has been happening is that the molecules of transistors started passing messages back and forth via electromagnetic fields, and selectively attached to each other to form an even powerful processor. The atoms not necessary to build the connection simply fell away. After those atoms were discarded, they were free to participate in the formation of a new connection. A lot of us here are very excited – we can’t even begin to speculate what it’s creating – this microchip is essentially creating itself atom-by- atom, molecule-b- molecule. We won’t really know what it intends to do for another few weeks. Algorithmically, the main challenge so far has been how individual atoms wouldn’t have many computational resources, but we expected this. The main question at this point is how we create economical algorithms that won’t waste information at such low levels of communication and at such low levels of storage.
The answer lies, as often is the case, in nature. You asked me what the ultimate goal was, so here it is: you want a neural and limbic learning machine that can sense its environment, is self–aware, and can plan for the future by formulating strategies. That’s it. What better hope for the future of our species than that? I’m talking about a self–replicating, self–guided, self–repairable, essentially indestructible learning machine with modular, unlimited information storage and processing capabilities, with a hardwired instinct to adapt to ever changing environments.
You may be thinking that
this sounds awfully familiar, and it is. But the truth is that we can’t simply design this thing from the top down. It’s too costly. Too much processing power is needed, too much heat would be generated, and too much cooling would need to be done, on top of the thing most likely ending up bulky and, for lack of a better word, dumb. Sure, we would have the fastest processing machine ever created, which could store obscene amounts of information. But it would just be a big, dumb, directionless machine incapable of understanding what it’s doing. We already have supercomputers like that – it’s nothing new. We don’t want another dumb machine that would simply carry out commands like some sort of complex Rube Goldberg device. Save those supercomputers for atom smashing. Again, we think that the answer can be found in nature. We think that this thing has to learn about its surroundings by trial and error, like an organism. One of the ideas floating around here lately is that we need to design this thing to be organic in nature, but mechanical in purpose – in short, we think it needs to evolve before it can one day be able to think like we do. The power to store and process information is rather meaningless, if whatever is doing the storing and processing doesn’t have a specific goal in mind. Mind, after all, is the fundamental issue. I’ll keep you posted on things as they develop.
Week Ending 3/28/35
Classified Restricted Ivory Tower Redshift
Lancelot Kattar, Center for Energetic Materials
Something truly amazing happened this week. We came into the labs on Monday morning to find that, over the course of the three-day weekend, the synthetic polymers threose and anhydrohexitol completely transcribed entire strands of XNA sequences. Without any outside introduction of existing DNA polymerases, our simple approach has proven effective, as our atomic transistors continue to evolve like their naturally occurring counterparts, DNA and RNA. This is amazing for two reasons: one, we programmed the initial algorithm to find and create the most optimal and efficient intermediary for trial–and–error, if–then clauses – this means that, without any influence on our part, our microprocessors automatically started creating a synthetic replica of the very building blocks for life! It’s as if our little microprocessor is creating its own probability clause, and coming to the same conclusion that nature did some three hundred million years ago. This is truly an amazing observation. The second reason is that this is absolute, irrefutable proof that you don’t have to adhere with the ribose or deoxyribose blueprints of RNA and DNA in order to have heritable, transferable information that evolves. It’s very strange – we don’t know where it’s getting its information. How this was supposed to work is that the polymerases synthetic XNA would get synthesized through rounds of DNA to XNA transcription, and vice versa – creating polymerases that can replicate XNA directly from XNA. But this new synthetic genetic sequence has done so on its own, despite both XNA strands being foreign to the polymerase – no DNA was ever introduced for this thing to model itself after. We have to look into it a bit more, but Alice and I are convinced – the only possibility is if some of our own DNA somehow contaminated the process without us knowing it. I mean, we thought it was really just going to create some sort of processor – some logic machine based on whatever interpretation it had for adaptation. We never expected this kind of evolution. It’s clear now that it wants to replicate life – after we programed that first microchip with the simple instruction to build itself so that it could be the most optimal adaptation machine, it immediately started to resemble life, just like how single cells act in nature. It’s amazing. Truly amazing. It’s building some new form of life that’s based on completely synthetic genetic polymers. We’re mere months away from witnessing the birth of a completely novel life form, that would have created itself out of the simple premise: adapt or die.
I had to sit down. Had we found some sort of synthetic life–form on Mars? I was getting the impression that whatever was being described in these reports wasn’t something that we found, but something that we created here in a lab. Maybe we used the artifact to make something else.
“Mo Stack,” I said. “Can you tell me anything about where on Mars the artifact was found?”
STIPULATE
“The artifact on Mars,” I repeated slowly. “Where was the cryptographic device on Mars found? The cube that contained nanotubes? The humming block of Martian regolith? The artifact that is currently sitting in the Clean Room at this very moment.”
SEARCHING
What happened? Suddenly there’s an artifact – I thought that there was always an artifact – but why was there no record of it being found in a chromite mine on Mars? Why didn’t Patrick know anything about it? That’s not entirely accurate, because he did know – he knew there was a device, an artifact – the part that he didn’t know was how we found it.
APOLOGIES, THERE WERE NO RESULTS FOR YOUR SEARCH
I sat back, remembering the file labeled excavation. I made a mental note to circle back to that later. I glanced at Sarah, who hadn’t taken her eyes away from the vault since she started watching it. The dials continued their slow revolutions in opposite directions.
For the time being, I decided to keep reading.
6.
Week Ending 7/3/36
Classified Restricted Ivory Tower Redshift
Lancelot Kattar, Center for Energetic Materials
Not much to report. With the growing need to get outside and participate in public health issues, I’m sorry to say that our attention has been diverted to more important things. Our little project keeps evolving so efficiently that it hardly needs our supervision at this point. I don’t know what it’s doing. It just keeps building, quietly running through its own little world. I should know more soon. Until then I’ll be in the streets, helping where I can.
Week Ending 8/13/36
Classified Restricted Ivory Tower Redshift
Lancelot Kattar, Center for Energetic Materials
We’re finding that the artifact has created what looks like some type of microtubule. These MT’s seem to connect these microchips by growing around the various points at which the separate devices connect. I wondered a bit how it was accomplishing the growth of new logic gates without lithography or etching… It’s also pulling molecules of carbon to coat the strings of amino acids, turning them into what look like microscopic wires. It’s amazing how keeping it simple is sometimes the best way to approach things. Who knew that a simple concept of evolution would lead to such feats of nanoengineering? What’s really interesting is that these MT’s demand less energy, one hundred times below those of traditional circuitry. Last thing, these microtubules also seem to know where to make an established link, because their telomeres have different polarities. We’ve had to start shipping in gallons of amino–acids and tons of carbon – both of which seem to be all the materials it needs to continue building itself. One more thing, Alice mentioned that the network of MT’s very much resembles neuron cells. I’ll keep you posted.
Week Ending 1/1/37
Classified Restricted Ivory Tower Redshift
Lancelot Kattar, Center for Energetic Materials
We couldn’t figure out where it was pulling its energy reserves from until Alice found an organic compound that was storing it. This compound also looks like it’s being organized in a way that resembles the membrane of cell walls. We’re not certain yet, and we don’t want to jump the gun here, but we’re fairly sure that this compound is also serving as a component to what looks like separate compounds that are being secreted between tubules. I have to say that I think it’s a type of synthetic lipid. I want to also say that this is remarkable, because the artifact would have had to learn at some point that this bio material it’s creating dies if it’s deposited directly onto the molecules of carbon. It must have figured that out on its own! It must have learned through adaptation to create a fatty substance that would protect the materials when they’re combined. I don’t think we are f
ully capable at this point to completely understand how amazing this is. The fact that we have a synthetic machine essentially mimicking the products of natural selection is just… too heavy to think about right now.
This circles back to my initial hypothesis. Most contemporary communication technology relies on electricity to carry the flow of data. Compare this to biological systems, when they follow a completely different paradigm that is entirely more efficient. Our bodies use a complex network of cell receptors, pathways and glands to control signal transmission – the algorithm of biophysiological reactions inside of cells – man–made devices cannot match biology in this context. Not even by a long shot.
Like I said – computers are basically dumb. You push a button and it generates sound and light – images pop up on a screen somewhere. Our goal at the onset of this project has always been to create a microprocessor, which has biomolecules that can talk directly to semiconductor logic–gates. The idea is to create something that thinks, that can feel something for what it does. The truth is that if we don’t reach that point, this whole endeavor is a waste. Might as well dump the Lexicon into a bundle of fiber optic lasers and fire it into the Kuiper Belt. We want something that could actively protect the Lexicon when we’re all gone – something that could care whether it and the Lexicon survive or not.
Week Ending 9/4/37
Classified Restricted Ivory Tower Redshift
Lancelot Kattar, Center for Energetic Materials
We had to relocate to the second floor to make room for a makeshift triage center. We worked remarkably fast, and I think the team had everything torn down and moved in under four hours. I think the news said that a total of three billion people have died so far of unknown causes. I hope all of your families are safe.
The artifact is still functioning, doing what it does. We checked in to see how it was progressing, and we were absolutely stunned. We have officially created the first ever synthetic learning machine, which has settled into a behavior of continuously rewiring itself every time it learns a new task through trial and error. This thing has minimal programing and absolutely no software at all. It’s turning into something resembling a neural network that is constantly rewriting itself, and there’s no doubt at this point that what we have is something that is following Hebb’s rule to the letter. When it gets something right, those neural pathways are reinforced – we have measured how it changes certain connections between these synthetic neurons every time a task is successfully performed. If for whatever reason its goals are blocked or thwarted, it simply pulls molecules of carbon, starts stringing together segments of amino acids in order to create what it needs to progress. It’s learning. It’s developing habits. It’s creating itself on the subatomic level the same way a child learns, by crashing into things. By having experiences. This is a completely separate architecture from simple digital computers. I mean, it still is that, but it has grown into so much more.