> Does AI result in greater productivity for engineers, and does greater productivity per person mean demand can be satisfied with fewer people?

Between the disagreements regarding performance metrics, the fact that AI will happily increase its own scope of work as well as facilitate increasing any task, sprint, or projects scope of work, and Jevons Paradox, the world may never know the answer to either of these questions.

There's a problem with the idea that hiring works efficiently. Twitter ran with thousands of engineers for a long time and clearly it did not need to.

It does improve productivity, just like a good IDE. But engineers didn't get replaced by IDEs and they haven't yet been replaced by AI.

By the time its good enough to replace actual engineers, any job done in front of a computer will be at risk. I'm hoping that will happen at the same time as AI embodiment in robots, then every job will be automated, not just computer based ones.

That's because of overhiring and other non-ai related reasons (i.e. Higher interest rates means less VC funding available).

In reality, getting AI to do actual human work, as of the moment, takes much more effort and cost than you get back in cost savings. These companies will claim they are using AI, even if its just a few engineers using Windsurf.

The companies claim AI is the reason they laid off engineers to make it look like they're innovating, not downsizing, which makes them look better in the eyes of investors and shareholders.

But not because of AI, they only use that as pretext for normal layoffs. Sometimes they also use it to hire cheaper workers fresh from school or a cheaper country, so just replacing expensive seniors.