CHIMPANZERO

“So what brings you in today?” she asks, the very composite image of a therapist. Young woman, probably a psychology major. Unaware of the replication crisis. Unresolved psychological troubles of her own brewing far below the conscious realm, yet evincing themselves in every uttered syllable.

“Oh, it’s a long story.”

“Don’t worry. We have time.”

For years, it was common knowledge – of the viral, sticky, mimetic kind – that chimps couldn’t be scientists. This was sufficiently long after the myths had dissipated that chimps couldn’t do anything a smart human could. Modern neural I/O had proven that chimp brains, while a fourth the size of their human counterparts, ran algorithms vastly more sophisticated. They could model social situations at a degree of detail that gave any politician with access to a zoo and a brain interface a distinct advantage. Game theorists proposed that it was the chimpanzee’s very cognitive lead that had held them back. Any given chimp could model the dynamics of its tribe well enough to introduce divisive memes into the fray, so no individual ideas won out. The development of language and high-parameter count brains apparently required many such steps of one simple idea out-replicating the rest, and thus the entire chimpanzee race was relegated to ebulliently but incoherently screeching at the top of their lungs to get their points across. It was suggested that their lowest common denominator methods of communication served as a form of defense against deviant mimetic actors, sustaining their steady state.

Chimps proved hard to integrate into the economy. Their inscrutable minds often drove them to smash apparatus without warning. Only by siphoning compute out of their subconscious was anything ever accomplished.

The impossibility of chimpanzee-aided science became apparent once Alfonso, a young male bonobo, was tasked as the principal investigator of an aerospace research lab. After doing a concept transfer of all human aerospace knowledge into his own mind (the crucial technological step enabling this was the construction of a transcoder algorithm between human and chimpanzee concept encodings), he was able to design superior aircraft engines and model rocket shafts than the lab had produced in decades. The scientists running the program were too quick to celebrate, as they soon found out. Deep in the depths of chimp minds lurked a vicious phallic hyperfixation. Alfonso’s mind wouldn’t let him design rockets or fuselages larger than his own shlong.

Freudian restraints such as this were found to be pervasive among the chimp pattern language. The image that took hold in the scientific consciousness was that of evolutionary forces placing a polythene bag over the chimp intellect and pulling it taut. This was interpreted, of course, in various ways.

II.

Those working on chimp-aided science mostly disbanded, and headed off to employ their talents more judiciously. Some protested. Some breathed a disgusted sigh of relief. And some proposed to wield gene-editing techniques to design chimps with airplane-sized dicks, though the dwindling research efforts in the area settled on figuring out how to make chimps think they possessed arbitrarily large Johnsons.

A method was proposed based off the basic premise that enabled chimp-aided science: deteriorated concept encodings. Sophisticated minds instinctively try to repair constituent parts of their ontologies when cracks and inconsistencies are found. Chimp repair algorithms, alike most of the rest of their cognitive processes, proved to be better and faster than human repair algorithms. Load a physically inconsistent plane engine onto the brain of a chimp, and it will be rebuilt better than before. Why not, it was asked, load the concept of a chimp’s own phallus onto his own mind, with inconsistencies that would imply the correct version was as large as it could be at the limit?

For various reasons this didn’t work as stated. Though an important innovation was made in the din of all the desperately run experiments. Instead of making actual deteriorations to the concept encodings, simply change the error correction tag to signal damage had already been done.

This technique was not explored sufficiently before the Alfonso experimental results became widely known. Pictures of chimps took the place of the archetypal beta-male figure in online discourse, and “Alfonso” became an epithet synonymous with “virgin”.

Research grants in all chimpanzee applications began to peter out as animal rights activists became convinced all chimps lived in perpetual adolescent agony, and demanded the creation of a new cause area.

Those in the field who were wise bolted for their university’s machine learning departments. Those who were not so wise knew in their hearts what Job knew in his heart whilst God brought ruin upon him. A golden future lay ahead. We would solve chimp alignment on all levels and transition the economy to a bonobo substrate we controlled.

III.

As an undergrad, I slavishly pushed papers and poured over cryptic datasets for tired postdocs working on human-to-chimp concept transcoding. I did it without complaint because I knew I was venerated by fate, blessed with a powerful mind, and born to gift humanity – if not born as a gift to humanity. I would strive to make my name in science, and the winds of destiny pointed at chimpanzees, and the ferrofluid of the future clung in brilliant spikes to the magnets that were their minds.

I read the first few - breakthrough - Alfonso papers voraciously, and stayed up late with giddy excitement as the pieces of something wonderful coagulated in my mind. We had solved the transcoding problem, and it implied two things: concepts could be noninvasively exfiltrated, and concepts could be noninvasively inserted. The concept of concepts knew no bounds… Could we encode the concept of the superior chimp encoding algorithm into a chimp mind? And could we point the algorithm at itself? If so, it implied we would soon be banking against the theoretical limits of compression - which meant banking on the theoretical limits of intelligence. It also implied we would be able to select which algorithms a mind used, meaning humans would unlock the full potential of our huge sapient brains. Our reign of carnal and cognitive superiority would soon resume.

With a fiery passion I thrashed myself against the issue, resolving countless problems, banging my head against compiler errors and issues with our wetware (I restate that chimps are not easy to wrangle). But I was stuck, juxtaposed between being too early, yet simultaneously too late. This was before the floor of funding had vanished beneath our feet — but also before we figured out we could tag concepts as damaged without actually damaging them. If you load a broken repair algorithm into a chimp brain, by definition it won’t be able to self-repair, and the chimp will enter cerebral stasis.

I scarfed down the error-tagging papers, growing distraught as chimp-aided science befell the fate of psychoanalysis before it.

We no longer had the money to run our code on the zoo’s chimpanzees, but I sneaked in a backdoor into a few of the chimp-brain interfaces, and resumed my work covertly in the zoo’s cafeteria. The chimp’s behavior wouldn’t undergo a noticeable change no matter what I did to the self-repair algorithm – as long as I restored to a backup every so often.

Months passed, and I ran out of cash. My personal fantasies had grown gray, and I had become a barista at the zoo’s cafe, stubbornly continuing my work, stubbornly refusing to join an AI startup as an overqualified engineer. I mourn humanity its lost dream time, snatched away by Alfonso’s chimp inhibitions.

IV.

My therapist cannot stop laughing. A PhD under a top researcher at a prestigious university turned suddenly worthless by internet memes. About chimp penises.

I convince myself it’s the flaring up of an unspeakable childhood trauma, and manage to feel sorry for her.

“Sorry,” she sputters. “That’s hilarious, you have to admit it.”

“Aren’t you supposed to ask me “how does that make you feel?” or “I’m sure there’s an upside to this situation if only you could see it” or something?”

“Look, it wouldn’t be funny if it weren’t for the fact that you have a million highly lucrative options at your fingertips. You’re smart. You know you’re smart. You know that you’re held back by your own obstinacy. So what’s the issue here? Just work on the chimp stuff in your free time!”

“I suppose so,” I mutter.

“Well, our hour is up. I’m sure you’ll turn everything around by the next time we meet.”

I bitterly leave her office.

Seven hours later, I sneak past the zoo’s closed gates, and into the cafe. A rather duplicitous thing, given my entry is enabled through a stolen ID card, but I know I’m close to a breakthrough.

The zoo cafeteria is cold, and the fluid in my joints ices over as I pull out my laptop and establish an SSH tunnel into the chimp implants. The hours pass in a flurry. The sun pokes over the horizon, and casts really long shadows of the lunch tables. My brain has become an adrenaline mush.

Over my tenure as a barista, I managed to patch almost every issue with recompiling algorithms on chimp brains. Throughout the night, I cracked the remaining bugs with error-tagging. Now all that was left was to point the repair algorithm on itself.

My fingers dance on the keyboard as I write out that final compilation command, and I feel the gates of a new era unlock before me.

Then, only slightly registering my whole body shivering in anticipation, I click the return key.

Nothing happens. It’s all flat lines on graphs, and quite anti-climactic. But all exponential growth starts out as anti-climactic. Time to call it a wrap, and head home. As I shut my laptop, I see the first of the encoder algorithm’s efficiency metrics tick ever-so-slightly upward.