arrsingh.com

Distributed Systems & Artificial Intelligence

All Hail the Software Engineer

I believe that the AI-Led Demise of the future Software Engineer has been greatly exaggerated. Its easy to fall into the conventional thinking trap that we'll need fewer engineers in the future because AI assistants make developers more productive. In my opinion time is going to show that conventional thinking is wrong.

I'm betting that over time, companies will hire more engineers, not fewer, because of the fundamental nature of software[1]. So while an organization might initially reduce headcount (because 10 engineers accelerated by AI are as effective as 100 unassisted), the long-term winners will recognize that 200 engineers can scale an organization to 20,000 effective engineers (assuming linear scaling for simplicity). Teams that understand this will grow from 100 to 200 engineers, rather than shrink to 10.

This will reignite the war for talent, modulo the requirement that engineers effectively leverage AI.

That being said, the problem of how 200+ engineers should be organized into small effective teams that work together as a well-oiled machine to deliver consistently against business outcomes will continue to be solved by a combination of leadership and natural intelligence.

See discussion on LinkedIn


[1]: Software businesses have no constraints - focused teams can always build more products, features, and fix more bugs - limited only by product & engineering bandwidth.

Sign up for updates

No Thanks

Great! Check your inbox and click the link to confirm your subscription.