Last week, Palantir CEO Alex Karp said something tech leaders usually avoid saying out loud: AI is about to knock college-educated women down and lift working-class men up. His claim is simple — AI will disrupt "humanities-trained — largely Democratic" workers (mostly women in white-collar roles) while making "vocationally trained, working-class, often male, voters" more economically powerful. The facts underneath seem genuinely real: women are overrepresented in admin support, clerical, and professional white-collar roles — all of which AI should replace. But whether this is a prediction, a problem, or something Karp is actually rooting for remains the question.
1. This Is a Class Correction, Not Misogyny (Alex Karp, Louis Mosley)
AI is fixing a generational imbalance where college credentials mattered more than hands-on skill.
Karp's frame is about power structures, not gender per se. The US spent 40+ years treating a college degree as the only respectable pathway to the middle class. This, in turn, created a two-tier system where college-educated workers accumulated economic and political leverage. Vocational training — which can lead to stable, well-paying jobs in construction, HVAC, plumbing, and electrical work — got stigmatized as "lesser." AI doesn't create this shift because it hates women; it creates it because white-collar cognitive work becomes cheaper while hands-on, hard-to-automate vocational skills stay scarce.
Louis Mosley, head of Palantir UK, makes the same argument: cutting away "lanyard class" bureaucrats actually empowers workers who've been locked out of credentialism for decades. This is repositioned as equity—a rebalancing after the last generation overweighted college-educated credentials.
2. This Is Misogyny Dressed in Economics (Jezebel, Will Bunch)
Karp is celebrating the diminishment of educated women's economic power and pretending it's just market mechanics.
The tone matters, and Karp's tone is celebratory. He doesn't say "we need to prepare for disruption" or "we should help displaced workers retrain." Critics including Jezebel and Will Bunch see this as Karp saying the quiet part out loud: tech wants a less educated, less questioning workforce, and they're betting on AI to deliver it.
Women aren't in high-risk roles because they're overeducated — they're there because of misogyny. Women have been concentrated in administrative support, customer service, and junior legal/accounting roles not because they "chose" humanities, but because men pushed them there. Knocking them down further isn't correction; it's compounding.
History offers a precedent. In the 1980s, computerization automated secretarial roles (90% women). Those women didn't retrain into something better. Many fell into lower-wage service work or left the workforce. Their wages never recovered.
3. Let's Correct For The (Very Real) Risk (Brookings, Noreena Hertz)
AI will definitely disrupt white-collar work and hit women harder. Whether they'll retrain or just fall down is an open question.
Office and administrative support has 18.8 million US workers, about 85% women. Women's jobs are roughly 3 times more likely to be automated than men's in high-income countries. White-collar occupations — law, accounting, software, information tech — are definitely going down. And these occupations face far fewer alternative roles if their job disappears. Karp isn't making this up.
The question is whether they'll retrain. Noreena Hertz doesn't think 21 million women in high-risk occupations are going to smoothly transition into plumbing and electrical work. Age matters (you don't retrain at 45 the way you do at 25). Geography matters. Skills matter — switching from accounting to HVAC isn't moving laterally, it's starting over. And adoption gaps suggest women might fall further behind: men use generative AI daily at work at 36% rates, women at 25%. If women are slower to adopt the tools they need to stay competitive, they face double jeopardy.
Where This Lands
Karp is describing something real — white-collar work is about to get cheaper, and jobs requiring hands-on skill will stay valuable. The data backs that up. But if he's celebrating reducing women's economic power — rather than worrying about actual people losing actual jobs — that sucks. On the other hand, his critics may be confusing a genuine class-and-gender moment with simple misogyny. The real question is whether we do anything about it in time.