This is also true for software production: we have made writing and deploying software so easy to do that the rate at which training of software developers to produce PRs has outpaced the ability for systems either human or automated to handle managing the introduction of risk in all this new code.
This talk [1] by by Mark Sherman of the Software Engineering Institute at CMU is about as dark as the summary of injuries Tim presents in his post. He mentions there are on the order of hundreds of thousands of CVE reports that there are neither time nor money to triage.
As a society, "injury" (defined broadly to include software bugs as well as knee damage) has always been accepted for certain benefits. At some point we decide that the injuries are worth the cost of progress.
We send astronauts to space knowing that we will kill about 1 in 30 of them. We let thousands die on roads so we can move around cities quickly.
Everything we enjoy about modern society has a certain amount of injury risk.
We are constantly saying skew the bugs, let's see what happens.
People conflate risk (likelihood of the event) and hazard (amount of harm if the event happens) and I think it degrades our conversations. The issue with software is that hazard has quite a large range, due to class attack (hitting all instances of something at once, like a software update poisoned with malware) and I don't think even software developers understand the scale of a worst case scenario, let alone most politicians.
I feel like the idea that society has accepted software bugs for some other benefit is questionable, given that software has only had an impact on most people's lives for a relatively small period of time in relation to the amount of time humans have studied philosophy and ethics.
Edit: also I'm discussing security bugs where there are asymmetries in the risk profiles and threat models of actors and agents.
This talk [1] by by Mark Sherman of the Software Engineering Institute at CMU is about as dark as the summary of injuries Tim presents in his post. He mentions there are on the order of hundreds of thousands of CVE reports that there are neither time nor money to triage.
[1] https://www.platformsecuritysummit.com/2019/speaker/sherman/