AI Hiring Tools Can Yield Racial Bias and Systemic Rejection

Well, well, well - this is a bit of a doozy. Not shocking or surprising, but a doozy nonetheless. This report from Stanford highlights quite a few things:

• AI hiring tools are helping folks break the (weakly enforced) law around hiring bias.
• Blackbox (proprietary) systems help shield companies from this bias-as-a-service.
• "Black applicants are the most likely to be adversely impacted: 30% of Black applicants apply to at least one position that demonstrates adverse impact against Black applicants." For tech companies, this number is definitely way higher.

I wrote about my frustration, despite having a tenured background, in reading a book on "tech resumes" and on the job hunt itself. I don't believe folks are entitled to a job - I do think that if we're not going to provide durable and reliable employment, that we need to have society provide a social net with all of this excess capital such that we don't end up criminalizing them for things that jobs provide like the means of paying for rent, buying food and at this point not being sick. That's the silent part of the idea of labor: we're not entitled to security but we are guaranteed social exile (by homelessness) at best and prison at worst when forced to make moves undesirable to the "job makers."

A bit extreme from the realm of tech workers - especially the highly paid ones like myself (perhaps when you're not floating two families) - but it's the truth of the system that we silently enable by not doing anything about it. I do not expect people whose job is largely to speak on the greatness of industry (Lenin was right about https://www.economist.com/, as a subscriber for two years now) to seriously address this. That's a problem because they become the canonical voice of industry and there's a class of peers - kinfolk or not - who use them as a barometer of how good the industry is going despite reported issues being held up elsewhere.