Will AI create or take away more job opportunities?

AI's net job impact is still unfolding — I'll pull recent research on where it's displacing and creating work.

On aggregate, most large-scale economic forecasts say AI will create more jobs than it displaces by 2030 - but it will displace a very large number of jobs at the same time, and whether workers benefit depends on reskilling.

This is the pattern all the major studies converge on:

1. The headline forecasts are net positive, with huge churn

In other words, the net is +78M globally, but 92M existing roles still disappear.

2. Exposure is much larger than net displacement

Exposure = jobs where AI could do part of the work. Loss = jobs actually eliminated. Forecasters stress the difference:

Crucially, both IMF and Goldman Sachs split exposure in half: for about half of exposed jobs AI will likely complement workers and raise productivity, for the other half it could substitute tasks and lower labor demand.

What that means for actual job loss:

The IMF is explicit: the 40% exposure "does not mean that all those positions will disappear."

3. Why workers are worried despite a net gain

The creation and displacement do not happen to the same people:

> "the jobs being destroyed and the jobs being created are not the same jobs. They don't require the same skills. They don't pay the same wages. And they aren't located in the same cities."

Near-term data shows the pain is already concentrated:

4. The new jobs look very different - AI-native

The WEF's other key finding is a wage and productivity premium for workers who use AI:

New categories that barely existed 3 years ago are posting triple-digit growth: AI Engineer roles grew 143% year-over-year, Prompt Engineer 136%, according to Autodesk's data cited by McKinsey. Demand for AI fluency rose from 1M to 7M US workers in two years.

Bottom line

The decisive variable cited in every report is reskilling: The WEF says 59% of the global workforce will need reskilling or upskilling by 2030, and 11% are projected to need training but not receive it. Without that, net creation at the macro level can still mean net job loss for individuals, regions and lower-wage occupations.

Historical precedent - from mechanization to computers - has been net job creation after a painful transition. Most economists, including McKinsey and IMF analysts, expect AI to follow the same path, but with faster skill change and deeper impact on white-collar work than previous waves.