I loved this post about engineers becoming managers for AI.
But I wouldn't say they were managing...
Here's the post: https://annievella.com/posts/the-software-engineering-identity-crisis
Professionally I haven't written any production code for about a decade and only small helper stuff since then. Otherwise I've been a manager, coach, consultant. I've often said that you can't be a true senior engineer unless you are mentoring juniors, and I've always expected mid & senior engineers to mentor and direct the work of junior & mid engineers.
With the people I managed and coached, the growth trajectory for senior (or lead or principle) engineer experience ended up looking like: pairing sessions, two way technical conversations with non-technical or engineering-adjacent, helping teams make good high-level decisions, making strategic and tactical decisions about how to solve problems ... and some coding on the side, even if just to make a proof-of-concept. Lots of engineers didn't enjoy that, though, and preferred to spend a lot of time writing code. (I called that "mid energy" but that's a story for another day.)
And now, thanks to AI, many engineers are doing that mentoring and directing, only with AI coding assistants!
I've called the best of what I do with people "managing". I wouldn't call this kind of work management, but I would call it "supervising". You're supervising someone else (real or virtual) in writing code and solving problems. Technical managers do do that (I did a bunch of that as an EM) but so does any engineer who has to lead and mentor.
The points of the blog post do still stand: how much do you want to be able to say "I wrote the code that solved that problem" vs "I explained the problem well enough to an LLM that it wrote the solution".
Good vibes everyone 🤙
