Exactly this. In any sort of survival-of-the-fittest environment, a redundant system can always be outcompeted by a highly optimised, brittle one, that dodges a few bullets.
Biology is one of the most ancient and cutthroat examples of "survival of the fittest" imaginable. Yet it's filled with highly redundant systems outcompeting better optimized and brittle systems. Cells spend huge amounts of resources on redundant DNA, transcription error checking, and redundant organelles. Animals have redundant organs, complicated immune systems, hugely expensive neural systems, and so on.
Human designed systems tend to be less redundant and comparatively fragile by contrast.
Very true! Though biological systems are also vulnerable to resources running out. Also, biological systems are frequently stupendously complex, as well as being robust (contrary to the article's claim).
At a guess, I'd think the difference is to do with the shape of the fitness landscape. If things are failing, or getting attacked and disabled all the time, robustness will result. Whereas a rare but serious failure is much less likely to be manageable.
Then again, perhaps evolution / god is just a better designer than us!
I imagine this can even be made mathematically precise. You need some minimal amount of robustness depending on the number of bullets you need to dodge, but conditional on that, someone who is just robust enough to be lucky will win, and if the population is large and diverse enough, a lucky, minimally robust, growth optimized individual will survive and win.
Nice! So excess robustness will cause failure through inefficiency (with "excess" being a situation dependent variable) - but you need some robustness. Evolution of course, optimizes for this, hence biological systems' wonderful balance.
I do think this is the part of the equation that is left out by the "precautionary principle" types.
