Good. My favourite is a human failsafe -- a Russian officer who refused to classify radar anomalies as an American ICBM launches, hence preventing WWIII.
Now I ask - are these really 'great towering achievements'? Or rather, are these just accounts of near-disasters narrowly averted by the failsafes that they sorely needed.
My point is simple - when the incremental risk is out of all proportion to incremental benefit, its best to scrap that technology.
In my book, that includes nuclear power (with the failsafes on offer now), nuclear weapons, and now... 'laser headlights' on cars.
incremental benefit = 30% off on the small fraction of gas which powers headlamps, doubling the range of headlights.
incremental risk: dazzling other drivers, blinding accidents (when lenses break), ubiquitous availability of technology that can be used to permanently blind large crowds of people