The Paradox Nobody's Talking About
Here's what should terrify you: AI exposure shows almost zero correlation with actual job losses. Yale Budget Lab and St. Louis Fed data from 2025 shows it clearly—76,440 to 77,999 tech positions were eliminated, but no occupational reallocation apocalypse. The correlation between AI exposure and unemployment? 0.47. That's weak. Statistically weak.
This doesn't mean AI is harmless. It means you've been measuring the wrong thing.
Exposure vs. Complementarity
The number that matters isn't your exposure score. It's complementarity—whether AI augments or replaces your work. Two consultants with identical exposure scores face completely different futures:
- High exposure + high complementarity: Uses AI tools to deliver faster, broader analysis. Becomes more valuable.
- High exposure + low complementarity: Gets outcompeted by AI systems doing the same work cheaper.
Your AIGMI score tells you how much AI can do your tasks. It doesn't tell you what your company will choose to do with that capability.
The Real Leverage Point
46% of jobs are affected at the task level, but actual displacement has been modest. Why? Because 92% of companies are still planning AI adoption—only 1% call themselves mature. Most organizations are still figuring out what complementarity means. This is your window.
The weakening correlation between exposure and unemployment gives you permission to stop catastrophizing and start strategizing. Your exposure is a capability description, not a termination notice.
What This Means for You
You have two levers: your exposure level (harder to change in the short term) and your company's complementarity choice (absolutely within your influence). Push for augmentation. Show how AI makes you better, not redundant. Document the 20% of your work that AI is terrible at. Make yourself the person who understands both the AI capability and the judgment call.
The job didn't change. The tools did. Whether you get replaced depends on whether you learn to use them.