Code and Other Laws of Cyberspace is a book by Lawrence Lessig, published in 1999. The revisions for the forthcoming second edition are being done publicly online at codebook.jot.com, with the enjoyable byproduct that I can read it for free.
It is one of those books which is superb despite being wrong. (Roger Penrose's The Emperor's New Mind is another). I will come back to it later, but there is one amusing metaphor, where the "Year 2000 problem" is described (in 1999, remember) as a code-based environmental disaster.
It's not clear, with hindsight, whether the "Y2K Problem" was (a) an over-hyped prophecy of doom, in response to which much money and effort was wasted, or (b) a genuine threat that was averted at the last moment at great cost.
My own professional experience, before Y2K and since, has been on systems that went wrong fairly frequently for all sorts of reasons, and were always fixed quickly, often by me. Of the dozens of production fixes I performed in the year 2000, only one of them was the result of a "Y2K" problem. The main (telecoms billing) system I worked on at the time was supposed to be replaced by a "Y2K compliant" one, but that project failed and the non-compliant system carried on without problems. I shouldn't really generalise from my experience to the whole industry, however, and I leave the question open.
The parallels with, say, Global Warming are numerous. It may well turn out to be a lot of fuss about nothing. Even if it's real, it might be most efficient to deal with the problems as they occur, rather than invest trillions in trying to prevent them in advance. A whole industry has emerged just to talk about it, and no organisation is allowed to omit paying lip-service, with varying degrees of insincerity, to the need to act.
Of course, analogies never prove anything. All this does is demonstrate a certain pattern of behaviour, which comes naturally to a human race raised on flood myths and the like. It doesn't mean the doom-mongers must be wrong, but at least it goes some way to refuting the "10,000 lemmings can't be wrong" part of the argument.