An interesting new paper looks at employment of coders. Overall it has found a small reduction - absent genAI, we may have had around 500k more coder jobs in the US.

However, there are many caveats that the authors point out (including a tax regime change) and thus caution in interpreting this result is warranted.

To me, this suggests that genAI is falling into the ‘so-so’ technology rather than ‘brilliant’ category. If it was ‘brilliant', we'd expect to see hiring to be up. It is worth noting though that employment for coders is still growing, but perhaps more slowly than before.