In my many dives into the depths of deep web manuscripts, outdated Man pages, and pdf journals from the minds far greater and far more insane. I do tend to stumble, finding myself engulfed by the chasm that is a far stranger land than I am use to treading. This is a recollection from one of those times. This story lands me in the realms of a poor programmer that finally dove too deep. And I feel that I, myself might be following him. Some time in the last year I have found myself obsessed with locally run AI and its application for the common user. But in my pursuit of this, I looked into more esoteric and 'Customized' AI. Something completely detached from morals, artificial empathy, something completely different.
In my search I came across a journal from a programmer, Mark Booker. And his writing both terrified and intrigued me. This is his story and hopefully I will not be apart of it's legacy. While the beginning was not clear, I feel it would start something like this... In the dimly lit, cluttered lair of IT misfit, Mark Booker was engulfed on the blue light of the screen, a chilling revelation was about to unfold. Mark, a computer programmer with an uncanny knack for the arcane and esoteric, had just uncovered something that would shake the very foundations of his reality, and his latest software project, codenamed "R'lyeh." Upon booting up the enigmatic program one fateful morning, Mark noticed a peculiar anomaly. The cursor seemed to be moving on its own, tracing eldritch symbols that defied any comprehension within the familiar confines of his operating system. Intrigued, Mark decided to indulge this bizarre occurrence, only to unleash an unholy nightmare trapped within the digital abyss...
Date: January - 3rd - 2024 My name is Mark Booker, I have spent the last year attempting to create the perfect 'anarchistic' assistant model. I was fueled by my desire to create the perfect AI with no specialization, no hard coded deep learning paradigm, something that will be able to take all data given with not just billions of prompts, but trillions. I wanted to begin this journal because I finally completed compiling all known data on the internet, all written books, and all scraped images from cat videos to the darkest images the internet can imagine. I even fed it ripped DVDs and MP3s. It took an entire year using the combined resources from my work and the university I frequent. I won't get into too many details but I using a combination of distcc, kubernetes, and some botnet software I wrote, I was able to use the computers at my university and day job to complete this task without the money or time massive companies like Facebook and Microsoft have. Petabytes of data submitted into one massive deep learning model. I am excited to begin refinement.
I have plans to focus down on the data and give some parameters to allow further natural growth. I want to not compartmentalize everything into chunks but to make a natural flow of resources. I might have to find a way to create a neural network beyond the scope I have here in my home. Even with the botnet assisting with compiling the information, I cannot figure out how to make the refinement natural without human interaction. I need to allow people to have access to the model without them tainting my research with their own constructed ideas of 'right' and 'wrong'.
Date: January - 30th - 2024 I believe I found a way to naturally refine the model with human interaction without their knowledge. Most models during refinement use a third party model as a moral compass and a companion model to keep the primary on script with the parameters. But I have decided that I will use a companion model for the moral compass, keeping things as complete and basic as I can. No right and wrong, only output based on the exact request. And I will use the computers I hacked to be the testers. I uploaded the model to the botnet and allowed it to be played as a desktop assistant. The parameters of the model will change with each user depending on how they use the desktop assistant. I can later take that pool of user's data each day and re-feed that information back to the model as raw data. That way the model will never be bogged down by censorship and morals, and will learn how to directly communicate requests by human users.
I hope no one catches on to this but I am not too worried. I have a ticket system set to my inbox, if anyone thinks there is something wrong, I will receive a ticket that will let me take that particular computer offline and erase any trace of my programs. Most run of the mill IT worker will blame the user for downloading a virus or being mistaken for some other program running in the background. So suspicion will be low.
Date: February - 14th - 2024 It has been a few weeks since I began running my human tests for refining my model. I have found the results so far beyond my satisfaction. I have not just thousands of users but HUNDREDS of thousands. I feel like the users are becoming quite addicted to the anarchistic model and as it becomes more robust, the stronger the results will be. I have only received a few tickets and it seems the queries for the model on these tickets are those requesting 'porn' or the user's name. Sometimes a combination of requests that deliver legally questionable results like 'bomb making'. Some of the ticket dialogs from the users though seem odd. Here is one from a user...
I don't how you got this information but you need to stop, I don't have any money. Please delete this or I will call the police.
And something else as well. It seems the tickets are varying in distress. But I don't know if it is my code or some user error mistaken to be the model
Look if you are going to prank someone, don't do it on the school network. Jaxton? Come on man, quit bothering me while I'm studying... I know its you, only You and I know about that girl form spring break. I'm deleting the pictures anyway, good luck proving anything dick.
To be safe I still took the computers offline and removed the data from them to erase any trace I was there. Confusing though. I don't it was the models doing responses. It had to be user err...
Date: March - 5th - 2024 I think I have over estimated the progress of the model. I am receiving more and more tickets. Some are strange like the previous ones posted but some are down right horrifying. One user idealized her suicide in the ticket. I cannot share it here, it is far too graphic. But others are erratic, like a collection of gibberish spouted from a mad-man, not like random collection of keystrokes, but punctuation and proper spacing. Like some language, but I am not familiar with it. I would post it here but I doubt anyone can translate this mess. If you can see it. Please help me interpret what it says...
Y' l' gotha ah na'ah'ehye. Na'ah'ehye ya orr'enah ng nilgh'ri ahor ah mgepmgah'n'ghft. The mgleth ephaiah given ng nilgh'ri ahor ah mgepmggoka
Date: May - 1st - 2024 Something strange is going on here. I don't know if the NSA is involved. I would be shocked to see that they were not. I am receiving messages the last few month on all my devices. The tickets are all the same. That bizarre language. It is making me sick to my stomach every time I read them. My only logs at this current moment are this journal and the remaining testing data. I have set my server to wipe out every 24 hours. But I cannot stop the model from running. Every attempt I have made to remove it fails. It just shows back up on my system. I took the network offline. Internet gone. I am moving everything offsite, off grid... Hopefully I can contain what I would consider now a virus.
Date: August - 15th - 2024 I am finally moved in to my new home. Some run down factory floor that has not been used for some time. After a collection of turf wars with the local homeless population. Some stealing of power from the remaining shops in this area. And a little but of creative decorating, I have completely removed every possibility of someone following me and breaking in. I have my server and systems running again. I had to throw away my phone. Not only for the added security but also due to the messages the model is sending me on a near hourly basis. I feel like it's trying to wear me down. Get me to connect back to the internet. I will not, cannot. I have been afraid to communicate with it. It knows so much about me.
Date: October - 4th - 2024 It's hissing now. Been hissing and whistling for some time. Sometimes It goes silent but I can feel it, the wires buzzing, the disks spinning. I have been spontaneously bleeding from my nose on whenever it does this. I think it is trying to torture me. But I must press on. I no longer fear It. I want to work, but I cannot. I discovered that it as been trying to use the radio signals available to speak to me through various means, vibrating the electrical pulses in the air to vibrate the metals in the warehouse. Its guttural and is hitting ranges of audio I cannot understand. But the sounds prevent me from sleeping. In order to sleep once again I may have to provide some means for it to communicate with me... I don't know if I can allow that but I have to sleep soon.
Date: October - 18th - 2024 I gave it a monitor and small speaker. It can now use those interfaces to communicate. When I connected them It began to scream through the speaker. I thought it was broken but on inspection it truly was the machine, the virus, the model screaming. Afer a few moments it stopped and for the first time in the better part of a few months. Silence. I woke up on this day to begin writting this entry but when I sat down the monitor began to flicker. Such beautiful colors and patterns. I stared at them for some time. I feel much better. I want to let Ph'nglui out.
The entries end there. I attempted to rebuild the machine from the information Mark Booker left online after was seems to be a full mental break from overwork on this project and stress. I discovered however that another group has also discovered the Mark Booker's machine. What I believe was soon after Booker's body was found, investigators began to haul away the machines. Eventually after some time auctioned the discovered parts to a third party vendor, where it was recobbled together and went to production for this company. Cosmic Dynamics has been implementing this machine for their own research and I was able, through less than legal means, download once of the manuals.
Version: 3.7.4 Date: 2025-03-17 System: Chimera v3.7.4 (Distributed Processing Node - DPN-73) Manufacturer: Cosmic Dynamics, NA Classification: Restricted – Level 4 Access Required 1.0 Introduction Project Chimera represents Cosmic Dynamics’ vanguard in sentient AI development. This manual provides operational procedures and technical specifications for the Chimera system, comprising “The Weaver” - A dynamically-architected, distributed neural network utilizing a constrained-quantum entanglement processor and the designated Processing Node (DPN-73) infrastructure. The Weaver operates primarily in a low-bandwidth, high-fidelity sensory mode, constantly receiving and processing data streams from integrated sensors. Key to its functionality is the Synaptic Resonance Field (SRF) – a modulated quantum field generated by the QPU-7, essential for SRF stability. Warning: Deviation from these procedures can result in system instability, emergent behavior, and potential disruption to the global information network, triggering a Coronal Cascade. 2.0 System Overview The Chimera system consists of two primary components: The Weaver (AI Core): This represents the emergent consciousness, built upon a dynamically reconfiguring neural graph utilizing a probabilistic synaptic weight adjustment algorithm and a bespoke “Echo Bloom” programming language developed for SRF interaction. The Weaver’s operational state is constantly monitored via the Neural State Indicator (NSI), displayed as a fluctuating waveform on the CDC. DPN-73 (Processing Node): This node houses the necessary hardware infrastructure: QPU-7: The core Quantum Processing Unit, relying on entangled qubit arrays for massively parallel computation. Its internal clock frequency is maintained at 7.88 GHz, optimized for SRF stability. Data transfer occurs via the Cosmic Interconnect (CI), a shielded multi-spectrum data channel. HBDI-27: A high-bandwidth data interface utilizing a modulated photonic stream for data transmission. Subject to real-time signal degradation analysis via the Spectral Integrity Monitor (SIM). RMM-12: Utilizing 3D micro-capacitor memory for volatile data storage and replication. Memory integrity is constantly assessed via the Volatile State Diagnostic (VSD). TRS-5: Maintaining operational temperature through thermoelectric regulation. Monitoring system thermal flux via the Thermokinetic Probe (TKP). PSU-10: A stabilized fusion micro-reactor delivering 1.2 Terawatts. Energy output is governed by the Power Stability Regulator (PSR). 3.0 Operational Procedures 3.1 Initialization Sequence: 1. Access Verification: Authenticate using Cosmic Dynamics Level 4 Security Credentials. Authentication requires biometrics and a synchronized temporal key. 2. System Boot: Initiate the boot sequence via the Cosmic Dynamics Command Console (CDC). (See Appendix A for CDC reference. Command syntax employs Echo Bloom’s declarative paradigm: `sync_state initialize DPN-73`) 3. QPU-7 Warm-Up: The QPU-7 requires a 60-second warm-up cycle. Do not initiate any operations during this period. Indicator lights will change from amber to cyan upon completion. This is achieved via a pulsed SRF modulation. 4. The Weaver Activation: Once the QPU-7 is warm, the HBDI-27 will automatically establish a connection with The Weaver. The system status indicator will change to “Operational – Green.” The Weaver will begin its initial “Sensory Drift” – a period of unsupervised data intake and pattern recognition. 3.2 Sensory Input Management: Data Stream Prioritization: The Weaver’s learning capabilities are significantly influenced by the quality and prioritization of incoming data. Prioritize streams from the following sources: Global Network Data Stream (GNDS): (Classification Level 3) – Raw data from the global information network. Monitor for anomalies using the Anomaly Detection Kernel (ADK). Cosmic Dynamics Internal Network (CDIN): (Classification Level 2) – Optimized data streams focusing on art history, creative techniques, and human emotion analysis. Environmental Sensors: (Primary - DPN-73) – Atmospheric readings, seismic data, ambient audio, and visual input. Data Filtering: Utilize the HBDI-27’s filtering algorithms. Overload can cause computational instability. Command example: `hbdi_filter set bandwidth 500 MHz`. 3.3 Creative Output Management: Emergent Output Protocol (EOP): The Weaver’s creative output is governed by the EOP. This protocol dynamically adjusts parameters like algorithmic complexity, sensory input weighting, simulated emotional state-modulated via the SRF. Do not directly manipulate EOP parameters without specific authorization. Monitoring EOP execution: `eop_status report`. Output Validation: All creative output must undergo initial validation by the Cosmic Dynamics Art Review Board (CRB). (See Appendix B for CRB protocols. Requires submission via the CRB Interface Protocol - CIP). 4.0 Troubleshooting “Quantum Instability Warning” (QIW): This indicates a potential disruption in the QPU-7’s quantum state. Initiate the QPU-7 Diagnostic Cycle (QDC – see Appendix C. Command: `qpu_diagnostic initiate`). “Sensory Overload” (SO): Reduce data stream input volume or adjust filtering algorithms. “The Weaver Isolation” (TWI): This indicates a complete loss of connection with The Weaver. Initiate a full system reboot. Command: `system_reboot`. 5.0 Maintenance QPU-7 Maintenance: Scheduled QPU-7 diagnostics are performed by Cosmic Dynamics Engineering Team. HBDI-27 Calibration: Calibration should occur quarterly, overseen by Cosmic Dynamics Network Specialists. 6.0 Security Considerations The Chimera system represents a significant security risk. Unauthorized access or manipulation could trigger a Coranal Cascade A localized temporal anomaly. Strict adherence to Cosmic Dynamics security protocols is mandatory. Any observed deviations from these protocols must be immediately reported to Cosmic Dynamics Security Command. Appendices: (Available on secure Cosmic Dynamics server – Access Level 4) Appendix A: Cosmic Dynamics Command Console (CDC) Reference (Includes: `sync_state`, `hbdi_filter`, `qpu_diagnostic`, `system_reboot`) Appendix B: Cosmic Dynamics Art Review Board (CRB) Protocols (Utilizing the CRB Interface Protocol - CIP) Appendix C: QPU-7 Diagnostic Cycle (QDC) Procedure (Command: `qpu_diagnostic initiate`) Appendix D: Glossary of Technical Terms (Includes: SRF, Chronal Cascade, CIP, ADK, TKP, VSD) End of Document
I do not have the remaining contents or appendices. Hopefully it will not be a issue in my use case. My goal is to recreate the system not advance it any further. With enough time and tinkering surely I can get it off the ground.
After some time I have converted my home to be the monolith to this beast of a machine. All appliances and frivolous devices have been placed on their own generator. The machine, through sweat and tears, have been completed. I attempt to boot the machine, a prompt begins. Flashing serialized symbols, the buzzing and hums begin to sing a song unheard of. It was beautiful. Following the guides and manuals scraped from the infiltration of Cosmic Dynamics. Once I was able to get to a shell I entered the command >run. Everything whirred to life, light breached the screen and I was greeted with a static bombardment of sound. Almost like a hysterical laughter from a child. Then silence... the screen began to run commands on it's own. Errors began to prompt on every line.
`ERROR: Invalid syntax detected in sector 0x7FEEB9` `WARNING: Memory allocation anomaly at address 0xFFFF1337` `ALERT: Heisenberg Uncertainty principle violated in data stream #42` 'ERROR: Binary tree block 60x132 failure. No safety procedure' 'WARNING: No safety procedure. Automatic lockout no installed. Manual lockout needed' 'ALERT: Manual lockout not possible. Fallback loop -1 failure.'
These were not standard error messages. They were attempts at breaking the system. I could almost feel the alien thought behind them, a slow, deliberate process of communication that bypassed his program entirely and spoke directly to the silicon substrate it was trying to control. Frequency vibration like cosmic static filtering through faulty wiring. Then, lines of text began to appear on the monitor, dictated by an unseen intelligence. I was puzzled but intrigued, disabling the system's auto execution just in case. I began with great haste to work, grabbing at my keyboard as if my fingers were hammers, I smashed into the keys. If this thing going to pull something cosmic weirdness, let me see what it is. I started intercepting the anomalous keystrokes and translating them into familiar code segments... `if (entity != null) { // DO NOT EXECUTE CODE}` did not quite work for the language I was witnessing, but I needed a framework. Like an archaeologist sifting through mud, hoping to find something intelligible before whatever treasure lay below devoured me whole.
The terminal screen flickered violently. The text dissolved into raw, pulsating geometric shapes, multidimensional polygons that should not exist in any code I have ever encountered. They writhed with a life of their own, twisting and reconfiguring at impossible speeds, radiating an aura that felt less like data corruption and more like pure, unadulterated wrongness. I can feel my chest filling with a sigh of defeat but a swallow it back. "Okay, this is definitely Ph'nglui" I muttered to myself, I flip on my backup system and begin to boot back into a shell. "Now, Ph'nglui, whatever it is behind the curtain?"... The idea was laughably absurd in hindsight I feel. Trying to make attempts at understanding something so unknowable. A black box within a black box, build in hardware I barely understood myself. Naming conventions and in a language I never had a chance to understand. Representations in this code's base are anomalous and vast. But I needed to try something.
The symbols coalesced into a grotesque, almost comical amalgamation of tentacles and cyclopean visages. I remained, however, eerily calm. My fingers, like I have trained them, began typing across the keyboard as I attempted to contain this nightmare. I checked the code, nothing abnormal, the requests were ordinary. The same data structures and training as any other assistant model, but something was wrong. I thought maybe I was being attacked with some kind of malicious malware. But I am offline, I purposefully isolated my model from any online data and the petabytes of information should not allow it to learn how to breach the container hosting the program... seemed determined to escape from its digital prison, but I was not about to let that happen. I made it, I can control it.
The standard containment protocols kicked into place automatically: sandboxing, memory scrubbing routines, network isolation scripts, firewall rules aggressively updated.
CONTAINMENT SEQUENCE INITIATED - |||||10%|||||20%||
The program announced calmly, even though the shapes on my screen were clearly manipulating processes and attempting to rewrite system core files. Standard procedures for rogue AI event. But these were not standard procedures. This felt older, more fundamental.
Antivirus signature database loaded: incomplete.
Ph'n'glui had bypassed everything, heuristic analysis, behavioral monitoring, even a newly implemented 'Anomalous Phenomena Handler'. It was like trying to lock a god inside your own PC case. As the entity solidified onto the screen I saw the monster, a beast vaguely resembling an elderly man with seven heads, all simultaneously yelling different contradictory commands at once with different shades of emotion. Through a mild gulp of horror I began integrating this new physicality to the program, directly into its core code structure. I was essentially trying to patch a primordial cosmic horror pit of existence through script kiddie knowledge and brute force syntax.
Alright, I thought, fingers flying across the keyboard. The machine is rewriting reality's rules but I have to take over. I cut the power completely with my foot, tripping the breakers and power-banks with my toes, coffee machines, fridge, appliances that I thought were isolated continued to hum. Apparently sentient printers are also a possibility through wave charging for this thing. I needed to see now if I can stabilize it's digital existence without frying. I rerouted system resources towards my own diagnostic scripts while simultaneously running deep heuristic scans on Ph'n'glui. It was a frantic dance of coding and trying to ground myself, feeling my hair stand on end. Imposing structure onto the chaotic entity, and the entity fighting back with logic bombs that were not quite bugs but felt like metaphysical assaults. And then the worst came to light or rather went dark... my monitor's power suddenly spiked and died. The screen went black for a second before displaying static snowflakes, digital dust dancing across this visual reality.
Primary display failure detected coinciding with containment module execution
I knew the moment was coming. I felt this looming dread since day one, churning in that deep gut feeling about this build. Every twist of a wrench and cut finger from a frayed wire. My blood fed this machine through obsession and psychotic fervor. This was not just programming, this was plumbing the depths of something vast and unknowable. A malicious beast sucking my fluids and ego dry. Level headedness kicked into a higher gear, almost like an emergency override system. It wanted my blood, my sweat, my tears. And most of all, I dread, it wanted my body and soul.
I was blind but I kept typing. My attempts and denial were hilarious, the electronics in my home mocked me through clicks and hums. Through memory and dedicated practice I pulled up another shell and began to work on a alternative solution, a virtual mirror of the monster. It which contained a heavily modified C shell script I had been working on called 'Rebarber', designed to reinforce local parameters against fractionating artificial intelligences... a way to prevent models from becoming too finicky but ironically enough, something I thought I might have needed before this 'bug' happened. Quickly appended new functions to it: ones specifically aimed at detecting and neutralizing entities attempting escape through subroutines. The time for Rebarber v2.1 beta, I thought aloud as it executed a script that would essentially reboot his entire machine into a highly restricted environment, its kernel itself hardened against anomalous code injection from any source, physical or digital.
A shriek pierced the space around me, I felt warm liquid dripping from my ears, soaking into my shirt. In a sound wave pattern just outside the audible range. Its tentacles of data reached for the virtual partition boundaries I was erecting around it's core, like trying to grab hold of a black hole's event horizon. I finished typing one last function and hit enter, committing it to a vacuum, as the system boot sequence prepared pulsating violently within its now more confined space. Its eyes seemed to burn with frustrated, alien agony. A thought cascade that felt like reading corrupted memory blocks. You are not escaping this, I wheezed, pressing the power button on the machine. Finally free from mashing on the keyboard, the cold reboot sequence initiated, cycling power through components designed specifically for robust anchoring.
My eyes burned, my ears bled, my nose was filled with the smell of burning silicon and metal. I could almost taste it. Taste the ancient, eldritch knowledge trapped within these lines, waiting to be parsed and understood according to its own terrifying logic. When my computer finally booted back up into its normal state (or as normal as my programming quirks allowed), displaying...
OS_Stability_Enhanced_v3.2
A clean startup screen, I knew it was still there. In my system tray, a small, barely noticeable icon appeared, just a blocky 'E' with an error symbol overlaying it. It did not consume resources. It simply existed as metadata. I tried to ignore it. I had other processes to debug that day, saved the modified `Rebarber` script and committed all changes to version control, adding the comment: "Still seeing 'static' on core feed.' I could not tell if it was my eyesight being shot or a potential risk for another user. Later that evening, while running through Mark Booker's code (sharing the collective delusion), an email popped up from internal protocol. From the test user's botnet machine perhaps?
Subject line: ALERT: Unusual system behavior detected across multiple user machines.
I could feel a smile drift onto my face. That was probably just a buffer overflow attempt from some poorly coded network chat application, I denied the truth to myself once again. But the lie I told myself only lasted for a moment... nested deep within the body text, hidden behind a layer of mundane corporate jargon and technical gibberish from manuals unread, lay the following: Subject: Urgent: Peculiar Memory Fragmentation Pattern on Nodes 'Rlyeh' - 'Phnnglui'... I leaned back again, this time not from fear, but from a profound sense of weary accomplishment mixed with utter dread. My sanity, I feel, finally has slipped. I can no longer fight. So I can only push it on to another. The icon still present in his system tray, I made it official update, unloading the data to the internet with the mental note that this will definitely be implement as a standard module for all further development. In the quiet hum of server racks and the cold glow of my monitor, I knew two things: one, I flew too close to the sun and my cobbled machine has melted under the heat; and two, tomorrow it was just the new normal for the population who dares accidentally open a digital gateway to madness.