Of course, we have to live with change. Things wear and need replacement (life cycle). Things become ineffective and need improvement.
As James Surowiecki reminds us, some innovative efforts have reaped many benefits (The New Yorker, 'Too clever by half?'). This technology, that is, the blog, is an example of an accumulation of a whole lot of progress. Were there missteps along the way (ACM is over 50 years old compared to the IEEE's 100)? Yes. Are MS Window crashes even countable?
Everything that is modern came from years of effort. Yet, is it always progress? That is, do we learn without failure (hopefully, with experimentation done under controlled conditions)?
And, if things are heavily model, and mathematical, in nature, then one expects that some type of verification step is involved. We don't expect to see drugs released without the appropriate scrutiny.
Yet, the marvels (in this sense, people) of finance have been allowed to experiment on the fly with our selves and our money.
Somehow, we have let those, who push their numeracy skills on the rest, to roam free of constraint, without requiring their due regard for reality (Quants, for example). Oh yes, the test (and measurement) was whether the efforts filled certain pockets.
Well, we have to get financial 'innovation' under control, somehow. Looking at a sandbox would a good start.
