One thing though is that the reason why there is an investigation in the Space X launch is that it vastly exceeded the estimate of possible damage. While no one was hurt directly, the cloud of debris from the pulverized launchpad apparently reached way further than projected, including inhabited areas. That means at best people having to clean their cars and windows (which is only an annoyance but still one that didn’t need to happen, though it could be easily fixed by Space X paying for the cleaning crews) and at worst health issues due to the dust and any possibly toxic components within it.
So that is, straight up, Space X underestimating a risk and underestimating second-order effects (such as, if your supposedly innocuous experimental launch that you flaunt openly is following a “fail fast and learn fast” methodology happens to cause trouble to people who have nothing to do with you, those people will be annoyed and will get back at you), so that the resulting mistake may indeed cost them way more than anticipated. Which is interesting in the framework of the analogy because while you probably can’t send an experimental rocket straight into orbit at first try, you probably also can at least do basic engineering to ensure it doesn’t blow up its own launchpad; this was simply deemed unnecessary in the name of iterating quicker and testing multiple uncertain things at once.
One thing though is that the reason why there is an investigation in the Space X launch is that it vastly exceeded the estimate of possible damage. While no one was hurt directly, the cloud of debris from the pulverized launchpad apparently reached way further than projected, including inhabited areas. That means at best people having to clean their cars and windows (which is only an annoyance but still one that didn’t need to happen, though it could be easily fixed by Space X paying for the cleaning crews) and at worst health issues due to the dust and any possibly toxic components within it.
So that is, straight up, Space X underestimating a risk and underestimating second-order effects (such as, if your supposedly innocuous experimental launch that you flaunt openly is following a “fail fast and learn fast” methodology happens to cause trouble to people who have nothing to do with you, those people will be annoyed and will get back at you), so that the resulting mistake may indeed cost them way more than anticipated. Which is interesting in the framework of the analogy because while you probably can’t send an experimental rocket straight into orbit at first try, you probably also can at least do basic engineering to ensure it doesn’t blow up its own launchpad; this was simply deemed unnecessary in the name of iterating quicker and testing multiple uncertain things at once.