Here's where AI is actually hitting the job market

Everyone's reading the labor data backwards. The "AI-resistant jobs pay more now" story is doing real work in career-advice columns — and the receipt people keep citing for it (the 56% premium from PwC) doesn't actually say what they think it says.

I went and pulled the payroll data. The trades premium is real, but the cause is boring — demographics plus the data-center buildout, not AI panic. The bigger story is on a different axis entirely. AI isn't moving the wage data yet. It's moving the hiring data. And the hiring tide moves first.

ℹ️ AI voice disclosure: This episode uses an AI-generated voice (ElevenLabs). The host is a pseudonymous byline persona (Reid Conlin).

In this memo

1. 0:00 Klarna's 700-jobs headline
2. 0:46 Pulling the receipts
3. 2:19 The 56% misread
4. 4:59 Hiring, not paychecks
5. 7:35 What Klarna actually wrote

Sources

• Brynjolfsson, Chandar, Chen — "Canaries in the Coal Mine," Stanford Digital Economy Lab (Nov 2025):
• Humlum & Vestergaard — "Large Language Models, Small Labor Market Effects," NBER Working Paper 33777 (Apr 2025):
• METR — "Measuring AI Ability to Complete Long Tasks" (Mar 2025):
• Anthropic Economic Index — January 2026 report:
• PwC — "The Fearless Future: 2025 Global AI Jobs Barometer" (May 2025):
• BLS Occupational Employment and Wage Statistics (May 2024, May 2025):
• Randstad — US Salary Guide + CEO interview, CNBC (May 2026):
• Brynjolfsson, Li, Raymond — "Generative AI at Work," NBER WP 31161 (2023):
• Indeed Hiring Lab — "AI at Work Report 2025" (Sep 2025):
• Klarna AI customer-service reversal:

Watch on YouTube ↗ Subscribe ↗