Some things age, some don’t. Some insights fade, some prevail. Inspiring 2015 talk by Dave Hahn about software engineering and operations at Netflix and in general.
Full of gems, but this one I liked in particular: Innovation is inherently error prone. It is called trial and error for a reason. So, whenever you innovate, you loose. You trade in a bit of stability, availability and “correctness” of your service. You can make that choice implicitly and hit those lazy engineers/devops/testers/architects/project managers over their whenever something goes wrong. Which will lead to people and teams that will avoid innovation and risk.
Or, you can go the way of Dave Hahn and make the trade explicit. Announce publicly that you do it and that it is okay to loose some stability and availability of your service for the sake of innovation. Then you fail. And deal with it the right way, by making failures unique. Ensuring instant reviews. Digging for the what and why in a save, open and honest environment. If you do that, you might arrive at the real root cause of what went wrong, fix it and move on.