Performative Engineering
Published ....
Last modified ....
From Boston, MA
Share this post on BlueskySee discussion on Bluesky
There’s been a decent amount of writing published as of late about the uptick in work that looks like work, but isn’t actual work.
I’ve started internalizing this concept as “performative engineering”.
This isn’t a net-new concept with AI and agents, but its definitely become more prevalent thanks to them.
When I talk about performative engineering, I mean things like:
- Announcing a new internal agent harness that you’ve spent the past week vibe coding
- Promoting a set of
markdown documentsskills to the#engineeringslack channel
Counter to what you might be thinking, I am a proponent for building better tooling, however I feel like a lot of engineers are optimizing for the output instead of the outcome thanks to AI and agents in general.
These explorations can be useful, and spending time to feel out the capabilities and whats possible is great, also sharing those learnings with others is a good thing too.
However the sense I’m starting to get is that leadership at companies will start to bias towards seeing these efforts as a signal that those individuals are shipping value to customers as opposed to improving the bottom line.
I don’t really know where we go from here. I honestly am pretty torn on what to optimize for at the moment, e.g. should I lean into this performative engineering trend to get visibility? Or risk being overlooked by focusing on the users that my team directly serves?