Chris Clearfield

I'm really interested in system failures, and have read a lot on the topic. Normal Accidents: Living with High-Risk Technologies, most notably. Meltdown is basically the pop-science beach-read version of Normal Accidents. It updates the topics a little bit and makes a convincing argument that most things have recently become more complex and tightly coupled, leading to more opportunities for massive system failures. (Although I will argue there is a huge difference between the Fukushima disaster and Starbucks doing something dumb on social media.)

The book is interesting, but it totally lost me when it started using sociology research and fMRI studies as the basis for exploring what can be done to help prevent systems failures. My eyes rolled nearly out of my head. That isn't to say that the author probably isn't correct -- diversity, outside perspectives, all of these things are important. But a study that looks at how people eat cookies after a meeting, that's not convincing to me, and it doesn't make me think that the author's answer is scientific at all.… (meer)

As the world has become more complex and connected, the conditions for meltdowns and disasters have become more common. As previously unconnected things become linked, they may interact in unpredictable ways; and as more things become connected with less flexibility for making changes or fixing mistakes, the mistakes that do occur have more serious consequences. In Meltdown, the authors illustrate how some catastrophic failures occur, but most importantly they explain how to mitigate those failures if they do happen or prevent them from happening at all.

I liked the positive approach to solving complex problems in this book and the fact that it was not limited to the big sort of complex systems — the solutions in this book can apply to that complex system known as “a family with a large number of people in it”. The book is clear, accessible, and compulsive reading. I’d recommend it if you read and enjoyed Careful, by Steve Casner, or if you find accident investigations interesting.… (meer)