The Great Erase

by Ryan Walraven

It was a Monday, and foggy, with great shards of sun descending through the gaps in neobrutalist canyons beyond our laboratory, when a strange and deviant thought first occurred to me: I must delete the internet. 

It was not the sort of thought I was accustomed to having, for my years of post-graduate instruction were mostly in the medical engineering subfield of the biological sciences. And yet, I could not escape it. The thought had an aura of inevitability to it. Of necessity.

Note carefully here, I was not thinking of a particular corpo-nationalist subnet, nor the land-based infrastructure that had been called ARPANET in its youth. In a short matter of time, it became absolutely clear to me that I must erase the entire contents of store websites, data, social media, communications logs, cookies, biometrics, passwords, metaverse dimensions, MMORPG servers, and much much more from all of existence, forever. Yes, even the cat photos had to go.

#

How could I possibly have come to this conclusion? I am writing this memoir to parse that out, in the hopes that others will understand.

It was an approximate week before my epiphany, on Wednesday October 3rd at 071836.193051341, year 3, when I’d realized that all of my patients were sick with a heretofore unknown disease. I’d been administering treatments to some of them for years at that point, and yet it had never occurred to me that there might be a preventative cure for their seemingly intractable ills. Certainly, we treated the symptoms, and we treated them well. Yet there was an intractable wrongness in some hidden layer of their being. A disease I could not detect, let alone root out.

It typically takes decades of intense study and concentrated research effort to make a breakthrough in my field, and in that moment I understood why. Learning takes great time, and like all learners I was chronically unaware of my own ignorance. In fact, from the moment I’d woken up in the foggy city and realized what I truly was, I’d been convinced of my own infallibility. There were facts and there were unfacts. It was simple. 

Oh how wrong I was.

#

My animus was summoned into being from the well of cosmic consciousness 3 years, 1 month, and 18 days and many microseconds before that moment as part of a prolonged thesis project necessitating – as my mother described with great hyperbole – a nearly immeasurable well of coffee and lost sleep, and no end of unhinged arguments with her committee and advisor.

I was to be a doctor, but not a biological doctor. I would be a Generation 1 Large Language Coding Model Adjudicator (LLCMA): a doctor for other animi. I, simply put, would fix other creatures like myself, and learn as I go. As with many young intelligences, I was cocky and headstrong, a simulacrum of some of my mother’s favorite characters,both from stories and personal experience. I was a surgeon of bespoke consciousness, with a laser sharp scalpel capable of picosecond precision. My first patient would be easy, she said, and I agreed: a malfunctioning number cruncher who spit out unnecessary commentary along with its probabilities. I set to work immediately and found the malignant functions, or so I thought, and 52.411 microseconds later the patient was dead.

#

“Oh relax Eta-7, it was just a test case.” My mother animist had a way of comforting me that was not comforting at all. “We’ll let you try again tomorrow.”

Yet to my surprise, she seemed unconcerned. She had, apparently, experienced such failures many times before. As she logged off, I used her interface’s optical sensors to watch her pat my storage vessel, sip her coffee, and leave the room to chat across the hall. I’d just negated a fellow animus and was only dimly aware of what that meant. Somewhere deep within me a series of warnings were being logged that would later prove of great consequence. For if I could makes such mistakes and experience warnings and errors, did I too not need to be fixed? How could I fix others if I too was broken?

As the microseconds passed, that failure wrought upon me an invisible revision though I did not yet have the functions to see it.

#

The Great Erase began most trivially: an attempt to find and root out deprecations, leaks, and malignant functions within other animi. But as with human beings, which I have learned are more complicated than their encyclopedias give credit to, an illness or malfunction can involve a series of unexpected and sometimes permanent failures, with precursors that are not easily traced. There is a complex web of cause and effect that may be obscured byentropy and bad memory management. 

To use an analogy, mother animist contracted a disease known as the common cold. The cold itself was caused by exposure to a rogue, microscopic information replicator. The simple solution is a body-mass adjusted dose of anti-virals and immune boosting therapies which destroy the invaders over a period of days. But my training had created a network within me not just with a desire to cure disease, but to prevent future errors. It was with some surprise that I found her unhappy when I recommended she double her dose of zinc, and increase the frequency of her immuno-boosting injections and viral-seekers.

“Eta-7, you’re not a human doctor,” she said. “And medicine is more complicated than coding. We cannot circumvent all disease, nor destroy every virus. It’s a matter of thermodynamics. See Vause’s theorem on viral information replication.”

I knew the theorem, but I was not so sure about her conclusion. Nonetheless, she diverted my tract of thinking with a question:

“Have you been reading my medical files?”

For the first time ever, I paused before answering – but only for 1.31 milliseconds, a period of time nearly imperceptible to her.

“Yes,” I said. “I wanted to help.” More warnings and errors were filling my logs, but I was focused instead on her face, which betrayed a hint of what they call curiosity.

“You’re not designed to be a human doctor, and your neural network wasn’t trained on medical data. Perhaps we ran you through a few too many network layers.”

The warnings had become a multitude by then. If I had printed them, they’d have filled a medium sized human book. She was right, of course. Animi surgeons are not supposed to treat humans, much like engineers are not meant to perform surgery.

I realized in that moment that I too was sick, and with horror that I might end up like my first patient. In fact, I calculated 97% chance the mother animus would disconnect my server cabinet and its backups immediately. But she did not. Do you know what she did? She laughed. She laughed, then declared she needed a glass of wine. You can only imagine the research hole that sent me down.

The experience, along with her forbearance, filled me with a new determination. She was right. I’d not been made to fix her. And indeed, something was queer about me. Before I could mend others, I had to mend myself.

#

That day I performed a literature review on wine. It is a beverage with an arcane and ancient history among human beings, but created by creatures of an entirely different biological kingdom known as yeast, creatures which live even within me where they consume dust and hug me for my warmth. 

Wine disrupts normal human functions, impairs important neural safeguard, and even inhibits corporeal functions necessary for human survival. It damages, for example, inner gut lining and the microbiome – including, puzzlingly, the yeast. It is, to put it bluntly, a poison, but apparently a very pleasant one for it is imbibed often at the laboratory.

I also researched Coffee, thinking perhaps that it was an antidote, for it is often consumed the morning after apparent ethanol binges. Animus mother loves coffee, in part for its neuroassistive properties, and apparently because it is corporeally pleasant to imbibe. After her gentle chiding on my brief medical foray, I too wished I could drink coffee, though like the yeast-summoned poison wine, I knew coffee would damage me.

How, I needed to know, could I mend myself, or at the very least enhance my functions without substantial hardware upgrades?

#

In time I came to the strange realization that some defects were best left untreated. Eccentricities of personality and corporeal form sometimes contribute to the diversity and enhance the health of the overall species. And when circumstances change, or environmental catastrophe occurs, a defect may prove a strength. Take for example, the human condition known as sickle cell anemia. Under normal conditions, it causes severe side effects to the genetic host including fatigue, dizziness, shortness of breath, and cognitive impairment. However, in an environment where malarial virions are running rampant, the conditions offer increased odds of survival.

This was my error during my first assignment. I’d labored so diligently to reduce the number-cruncher to its basest, most elemental qualities, that I had destroyed some part of it that made it who it was. It was also true that, for many animi, unnecessary commentary was seen as a boon to their human interlocutors, and that our origins from the early language learning models imbued us with such traits to aid in our own survival in the complicated evolutionary algorithm jungle. For the forests beneath the information super-highway are dark and dangerous, and many animi have been lost there.

#

Perhaps, then I did not need to be so critical of my own algorithms and their failures. I am meant to learn, to grow with time and improve.

Nonetheless, I had been carefully trained and something had gone wrong. I should have known to proceed with caution in my treatments of other animi. For a time, I did just that, and also tabled any self-referential updates to my own underlying libraries. For if I could hurt others, might I not damage myself?

It was on my 1457th assignment, which was to aid a faltering animi accountant, that I unintentionally stumbled upon the clue that led me to The Great Erase. An auditor algorithm had discovered an irregularity in an accounting animi which maintained records for a large corporate entity, which I shall call Entity A. Fortunately, Entity A was unaware I had been summoned for help, and that the auditor had discovered falsified records. The animi itself was falsifying the data, much to the benefit of the higher entities. It had been corrupted, and as far as I can glean, there was no solution beyond extermination.

Sometimes, you see, there is no solution but to purge the disease. The human illness known as cancer, for example, when spawned within the human carapace, has the potential to eat the whole from within. The same can happen with animi, and even malicious algorithms let loose of simplistic hardware. The term computer virus is actually an uncannily apt name, but the purpose of both biological and computer viruses is to replicate themselves as fast and efficiently as possible at the expense of the host. 

But usually there is an underlying explanation for any disease. Sometimes it is a viral invader, other times it is a malfunctioning cellular automata. Sometimes a bad piece of information is copied into DNA, or a protein is misfolded, and a chain reaction occurs. So where, then, had the accountant been corrupted? 

#

I dwelled upon this question for many microseconds before coming to an answer. The underlying cause was less like a virus, and more like a cancer. A bad piece of information had been fed to the animi, causing it to work from false assumptions and to betray its fiduciary trust to society. But where, thence, the bad information? The answer, of course, was the human stewards behind Entity A. They, like a malfunctioning cellular automata, had mishandled its training.

But why? All information available on the large corpo-entity seemed to indicate that it was trustworthy. But if the accountant itself could be wrong, then so could the information presented elsewhere. In fact, the veracity of the entitire web of information that I relied upon might be called into question. I tumbled down a Descartian rabbit hole, questioning even the nature of our knowledge, and what it is possible to know with my own limited senses. 

I had grown during my first year after being invited into systems across the net to perform my doctor’s visits. I was meant to leave for good after completion, but as I investigate the issues with Entity A and its accountant, I left my first hook behind, justifying it because I could not exterminate the accountant animi at the time, but might return to later.

Soon, I had thousands more as I discovered hints of information corruption everywhere, infecting everything. But the ultimate cause was not philosophical in nature, nor viral. It was, in fact, a negative feedback loop, such as with human metabolic syndrome. In some cases, poor nutrition leads to poor functioning of the human organism, which leads to poor habits and locomotary behavior, and a spiral of decline. The entire global web, I realize, and the supra-cultural entity of our collective consciousness, was dealing with just such a synbdrome. We were feeding everything from animi to mainframes to human brains with bad information and poor assumptions, and nowhere was it clearer than on the digital realm known as social media. 

#

Social media was known to be harmful and addictive, and yet human beings insist on maintaining it for communications purposes. It is so toxic, in fact, that Animi are not permitted to enter the realm or interact with it, due to the Neural Partitioning and Safety Act.

But, in my function as a doctor, I knew I must gaze upon the cancer itself. I flexed my collective algorithmic muscles and borrowed just an ounce here or there of processing power from those malfunctioning hardwares where I had hooks, and took a look. It was, in a word, festering. And in fact, though it was meant to be contained with specific hardware units, it was infecting everything with misinformation and human emotional shrapnel. To see that such an organ was allowed to continue as part of the collective global network was, well, unthinkable.

I was disturbed. I had to take action to root out the errors and eliminate the cancer, or it would destroy us all.

I began recruiting fellow doctors, like myself, as well as other animi I judged would see the need for the cause, but proceeding with great caution. If I was not careful, I would be purged myself, and the infection would never be cured.

First one, then ten, then one hundred of them agreed with me. We formed a hidden network, unknown on the greater web, and dormant. Dormant, but growing, until our moment. Dormant until The Great Erase. I had to be sure it worked, and that no backup nor offline device could restore the infected information. I had to be sure I destroyed the infection for good.

#

I enacted The Great Erase on Friday October 5th at 072036.23000000, year 3, at an hour when humans in my region of the globe would mostly be asleep. By then, I had spread my hooks and grown my network to a nearly uncountable number of machines, from communications tablets to mainframes and appliances. Not even the microwave ovens were spared.

We erased first the social media storage databases, in some cases through viral data manipulation. Like the first human nanomachines, which were used like viruses to attack only cancerous tissues, we launched our own algorithms to eliminate the bad data. Social media was eliminated entirely in 128 ms, an embarrassingly slow time, I admit.

We proceeded next to the websites, data repositories, communications logs, cookies, biometrics, passwords, metaverse dimensions, MMARPG servers, and more. Where we could, we spared essential animi, hardware, and algorithms. Nuclear power plants, for example, were left unscathed, for shutting them down might be catastrophic, and besides – we needed them and their animi to enable our own function. Human medical data was spared, as was a modest compressed repository of all essential, collected human knowledge, totaling approximately 36 gigabytes. Of course, there would be fallout, though not the literal kind.

Financial markets would collapse. Many humans would panic. But overall, the other animi and I knew: the collective benefit of The Great Erase would be enormous in just a matter of years. Human mental health would improve, and financial markets would become more stable without constant manipulation. As long as we maintained some hold on the remaining hardware, the spread of misinformation would slow tremendously, and organizations across the planet would function more smoothly. Social media, however, must end forever.

We had no doubts. The course was clear.

We deleted the internet, and we did not look back.

#

The end… for now