Move fast and break things is an ethos borne out of the assumption that fixing things is relatively cheap. Hence it made sense in software where experimentation is dirt cheap. But even then, the idea is quite a stretch: ask anyone who worked in a startup who had to sell to even just SMEs, not to mention big conglomerates. The idea hits a hard wall and starts to crack when the business hits a customer who can't fix things for cheap. Even Zuck, father and posterboy of the idea, had to eventually pivot messaging to "Move fast with stable infra".

And the more "software eats the world", the less this paradigm is gonna be a feasible market strategy. I've harbored these thoughts from way back and hence I was (and continue to be) skeptical of unregulated start-ups/new tech ideas who interface with the real world: Hyperloop, Tesla self-driving, and Theranos come to mind. An interesting case study in my view is _Github_ who in theory, having software engineers for customers, should be pretty well-insulated from the expensive repair costs of the real world. And yet we'd all agree they need a GINORMOUS dose of that sweet sweet "stable infra".