I find out what a CI is two days later on Friday, when we’re called to stand in a circle around the centre of the main hall, all fifty or sixty of us from both floors of the Group building. Simon’s cheery demeanour has evaporated, and he looks like he’s about to attend a funeral. There’s a hoarse manner in the way he fidgets with his hands.
“Right. Continuous integration.” His voice also sounds hoarse, like he hasn’t been sleeping. I can’t see his eyes to tell for sure. “Let’s start with the conversion enclave.”
Chang-dol steps forward from his place next me, confident as ever. “We have 10 new invites ready based on culture fit discussions, a couple flashy undergrads from MIT. Syrian and Myanmar outreach orgs getting record donations. Facing some difficulties with the GOF in Syria but we’ll manage to keep up an inflow based on current projections.”
“And the GQ system? Objectivity in our hiring decisions is important.” Something tells me Simon already knows what he needs to know about the current state of the GQ system. Eyes turn in the far ends of the circle. Clearly this has been the subject of public gossip.
Simon snaps his head around. “Alice? Why don’t you take over this section, I’m not as, ah, familiar with machine learning work.”
There is a moment of panic in Alice’s face, then she steels herself and takes a step out. “Based on comparisons with our hiring decisions, the model makes hiring decisions with 75% or above accuracy.” For a moment my heart soars, but then she continues. “Yet based on a group of randomly generated test applicants, the biggest correlation factors for a hiring decision are… education at elite institutions and… race. As for the GQ ranking, there’s too much variance from the different test metrics to get reliable results.” The last sentence comes out as a blurted mess.
There’s a wrinkle, a frown, and Simon shifts. His coarse voice raises slightly. “That’s not good enough then, isn’t it? More work must be done to find the hidden correlates that determine a good hire. Racism is the worst sign of eusocialisation of all.” Alice’s head is hung, flushed. For a long moment we all stare at her, then she breaks.
“Yes. My latent eusocialisation has infected my work. I apologise on behalf of my enclave and myself.”
To my surprise, Chang-dol also starts speaking. “It was my fault, for not guiding my team members correctly. My latent eusocialisation has infected my work. I apologise on behalf of my enclave and myself.” There is no trace of irony or resentment in his tone either.
At this Simon smiles. “Taking responsibility is good. You are too harsh on yourself.” He then looks at me. “We’ll skip you for this CI, William, but you’ll need to participate from the next one onwards. We have a culture of radical openness and responsibility in Radix, and everyone needs to play their part.”
And so it continues. One by one, team members step up and admit to their failings and laziness over the past week, making progress reports laced with self-deprecation. Simon doesn’t make many more comments, but when he does it always triggers a moment of intense self-reflection. At the end, even he has his moment of collapse.
“I apologise,” he says quietly. “I have been too absorbed in abstract academic matters this week. There was a paper on new transformer performance optimisations that absorbed my attention away from my more important work here. As a result, we have allowed oppositional metaorganisations to claim a larger part of the psychopower distribution than usual. My latent eusocialisation has infected my work. I apologise on behalf of Radix and myself.” He, too, seems completely genuine. Even when Simon dismisses the meeting and we update our goals for the next week, data still streams in to his glasses, obscuring his eyes from my vantage point as he returns to his office. For the first time, I notice that Simon walks with a slight limp.