Thought experiments are fun, but I am coming to a different conclusion. I agree that it changes the risk analysis, but I don't think that it's necessarily for t

Thought experiments are fun, but I am coming to a different conclusion. I agree that it changes the risk analysis, but I don't think that it's necessarily for the worse. Performance doesn't necessarily improve under pressure, and extreme risk aversion can cause some funny side effects - for instance working around known bugs rather than risk correcting them. Is there any data to support the "only possible result" that you conclude? Or, maybe examples of high-quality software that doesn't have any kind of high-frequency automatic update mechanism?