Once the legal niceties are concluded I am assigned to one of the many co-equal cooperative units (“enclaves”) in the Radix Group. My group is responsible for what in any other context would be called hiring or talent acquisition, but within Radix is called “conversion”. What it amounts to is designing a great battery of tests, some obvious and some embedded in games or other disguises like the one I encountered, to detect “steganosophonts” (also called “sophies”, “sophs”, or hidden talents) embedded around the country and the world. These people would be weeded out through our tests and then “elected” (sometimes “airlifted”, if the process involves a lot of paperwork or expenditure) to the Radix Group’s UK headquarters or one of their affiliated branches around the world. The plan, insofar as I understand it, is to collect a critical mass of these people such that they can convince the leaders of the world to adopt radical policy changes through their intelligence and expertise. I consider it a sort of hyper-ambitious think tank, and think little more of it.
As it turns out, the girl I nearly bumped into earlier is also part of the conversion group. Alice quickly and curtly explains that her goal is to rank applicants from these diverse metrics using a unified metric—what they half-jokingly refer to as “GQ” or “Genius Quotient”—by constructing a machine learning model to compare the data from these tests with data collected by existing Radix Group members.
“My current approach is based on anomaly detection with deep learning, we use an variational autoencoder trained on the result profiles of all current hires by the Group. The reconstruction error from this autoencoder becomes the metric to class potential applicants as inside or outside of a cut-off after feeding in their result profile, so applicants sufficiently similar to existing Group members over a range of metrics are considered—nevermind.” She sighs, then pulls out a slab of note paper seemingly from nowhere and tears off a sheet. She draws two overlapping circles, one labelled GENERAL POPULATION and one labelled GROUP HIRES. “It’s like a Venn diagram, anyone that overlaps enough with a generalised existing employee profile gets selected, and then we rank by various neurobiological metrics to get GQ.” She dots the intersecting zone a few times with her pen.
I take a second to chew on this piece of information. “Doesn’t that mean that the Group hires a lot of similar people?”
“Well, Simon double majored in psychology and computer science, and he chose the first crop by hand from his buddies if I had to guess. Since then hiring’s mostly the result of ad-hoc discussions about “fit”. But we’re mostly STEM people, yeah. That’s why he was so insistent on having alternative tests, like the one that you evidently got hired from.” She chuckles, not unkindly. “No offence, but I could see your eyes glazing over. I mean, I’m also a new hire. Some of these days I barely know what I’m doing either.” She notices that she’s overcorrecting and becomes quiet again.
“I mean, in theory there’s no seniority here, right? Simon’s just first amongst equals, or so he keeps telling me.”
She shrugs. “I mean, I guess he likes the sound of that. But it’s clear who calls the shots. I certainly can’t just spend thousands to start a new project on a whim.” A pause. “Mind you, I do get a lot of leeway to do what I want. It’s just… there’s a pecking order, one that isn’t written down.”
“Fair enough.”
She cranes her neck. “I think Changdol’s coming over. Gotta bounce.” She pronounces it as one word, Changdol. And just like that she’s gone.
The voice that greets me is also not unkind. “How’s the geographical qualitative going?” Lee Chang-dol is one of the old guard, the original hires brought on board by Simon, and also one of the few other quasi-humanists in the building. My job is to produce a list of reports (he insists on calling them “qualitatives” as opposed to the “quantitative” data analytics) on conflict zones and areas of instability for him, areas where sophs might show themselves and be easily recruited. Like Simon, he insists on wearing a suit to work, a sharp cut cream and tan affair that combined with his well-combed, greying hair makes him seem naturally, effortlessly imposing. I try to work up whatever jargon I can remember.
“The Syrian conversion team’s reports are quite optimistic, lots of engaged young folks with high non-eusocial tendencies, at least one localised potential soph cluster who scored above the cutoff on organising and management capabilities.” Eusocial is our nice way of saying “conformist, uncreative, generally passive”. Definitely not soph material.
“Mm, mmm.” He flicks his hand in a distinctly Simon-like way. “Continue.”
“A lot of applicants also report being approached by another group—” all of a sudden I can feel his gaze on my back “—some kind of optimisation based charity?”
“The Global Optimisation Foundation.” There is a tense edge to his voice.
“Uh… yeah. Those people.”
“What’s the percentage?”
“About 80%. Based on conversations with those approached they mostly report positive experiences—”
“Ssi-bal!” He spins around. “I’m going to update opops. Give me the rest of the quali later.” I’m still not used to this culture of rapid exits. Alice is nowhere to be seen, so I pop her a message.
W: What’s opops?
A: Opops?
W: Changdol just said he’s going to update opops and left.
A: Opops or op ops?
W: I dunno
A: So op ops. Opposition Operations. Theyre the branch that monitors rival metaorgs
W: Metaorgs?
A: Rival groups like us. Same plan, different goals. We fight for the same sophie population.
W: Seems… conspiratorial.
A: You’ll get it when we do the CI.
W: CI?
A: Gtg, you’ll know when it happens.