05/30/2026
From a draft of that article I have been encouraged to write:
In one recent application process, [COMPANY NAME REDACTED] advanced candidates through an AI-assisted screening funnel into a second round of prerecorded video questions.
Candidates were asked to perform spontaneity into a camera, answering broad behavioral prompts delivered by recorded employees with the mid-grade production quality of an 90's internal training module.
Before participating, I had to click a box to acknowledge that my video would not be used to discriminate against me. A quick search of the vendor’s recent history showed why that box existed. [VENDOR NAME REDACTED], a major video-interviewing platform, has been named in civil rights proceedings alleging that its product contributed to discriminatory outcomes.
The reassurance was almost too perfect: a legal checkbox placed in front of a process whose central promise is that machines can make human evaluation cleaner, faster, and fairer than humans can.
Then came the human layer, which was somehow stranger.
The prerecorded questioners in the process appeared carefully selected to project diversity and inclusion: women of color, professionally framed, earnestly asking open ended questions into the void.
The effect was not sinister so much as brutally comic: a company apparently responding to the anxieties of algorithmic bias with a video tableau that seemed to announce, “Look, humans work here. Diverse humans. Please notice. And don't sue.”
That is the whole white collar labor market in miniature: algorithmic judgment wrapped in human theater, legal caution wrapped in corporate optimism, and applicants asked to prove their humanity to a machine that has already shaped the terms of the encounter.