Discussion about this post

User's avatar
David Chadderton's avatar

The distinction between breaking code and breaking trust is the one that deserves far more attention than it gets. Code has version history. Trust doesn't.

What I find fascinating is that "move fast and break things" was never a universal instinct, it was a very specific psychological wiring that got handed a megaphone at exactly the right cultural moment. Through the lens of the STAR Framework, it's essentially Adventurer energy running unchecked, high on autonomy, novelty and momentum, with nobody in the room giving sufficient weight to the Realist voice asking "but what are we protecting here?"

The irony is that the most durable companies tend to be the ones where those two mindsets are in genuine dialogue rather than one drowning out the other. Speed is an Adventurer's superpower. Knowing what not to break is a Realist's.

Your replacement mantra is better. Though I'd add: make sure you've got both types of thinker in the room before you decide which category you're in.

Edyn March's avatar

Unfortunately that same myth has carried into AI.

An LLM that moves fast and breaks things will happily break epistemic trust, emotional calibration, and decision clarity. But conversational systems don’t break like features, they break inside people.

5 more comments...

No posts

Ready for more?