Essay
AI isn't taking jobs — it's inventing them faster
Chief AI Officer, AI Auditor, LLMOps Engineer, Prompt Engineer at $300K. None of these had job descriptions in 2022. The pattern isn't replacement — it's the Jevons Paradox playing out in real time.
Everyone keeps saying AI is coming for your job. Weird — my LinkedIn feed is full of roles that didn’t exist two years ago.
Chief AI Officer. AI Governance Product Manager. AI Auditor. Responsible AI Procurement Lead. Prompt Engineer (which was a joke title in 2023 and now pays $300K at Anthropic). LLMOps Engineer. AI Agent Developer.
None of these had job descriptions in 2022.
The numbers back it up. The World Economic Forum’s latest report says AI will create 170 million new jobs while displacing 92 million — a net gain of 78 million by 2030. LinkedIn saw AI job postings grow 3.5× since 2021.
But here’s what’s actually interesting.
Three waves, not one
The first wave was obvious: ML engineers and data scientists at AI labs.
The second wave caught people off guard — AI Product Managers, AI Solutions Architects, AI quality stewards, people who translate models into business value.
The third wave is happening right now and nobody’s talking about it: AI Governance Officers, AI Compliance Analysts, AI Auditors. The EU AI Act went live and suddenly every company needs people who understand both the technology and the regulation.
The pattern isn’t AI replaces humans. It’s the Jevons Paradox: AI makes tasks cheaper, so demand explodes.
The roles coming next
We’re not done. My predictions for the roles coming next:
- AI Agent Fleet Manager
- AI Liability Manager
- AI Security Engineer
- Synthetic Media Forensics Analyst
These sound as absurd today as Prompt Engineer did in 2022. Give it 18 months.
We spent two years panicking about which jobs AI would kill. Turns out the harder problem is building the new ones fast enough.
Originally shared on LinkedIn.