There are many approaches to writing a non-fiction book. One is to pick a thread that seems to have enough to write about, and then collect material related to that topical thread in different domains and put them together. In this approach there may not be a clear dovetailing build up from one chapter to the next, and so you can even read most chapters out of order, and it will make sense. The underlying thread certainly will remain visible, providing an overarching view. I did that to some extent in my own Thought Experiments book. Matt Parker has done the same and has managed to string together chapters that are quite accessible.
Book starts off with Leonard Vs. PepsiCo case as a clear example of how even TV commercials put together for big companies meant to be widely seen don't spend the time to check their math. It is an interesting case worth reading about, if you haven't heard of it before. Then he talks about UK lottery, where public had a hard time understanding -8 is lower than -6 and had to stop a particular lottery offering since many were complaining that though they "won" by getting a "bigger" number, they are not getting the prize! Next couple of chapters discuss errors in calendars, Y2K issue introduced into computing when years were coded using two digits, and a bunch of examples related to Microsoft Excel messing things up, when used in certain ways.
While these initial stories are funny, subsequent ones related to errors in engineering and medicine were actually deadly. There are stories about X-ray machine code that had math errors that allowed high dosage of radiation that actually killed a patient, bridge construction errors that resulted in dozens of workers dying in early 20th century, drug dosage calculation errors that had affected real patients adversely and so on. I had heard of some of the stories, such as the one about 20 Fenchurch street building in London that formed a lens focusing sun's rays heating up the area nearby dangerously. See the solar glare problem section of that Wikipedia page! There is whole section dedicated to blunders committed due to incorrect units usage (as in metric units Vs. British units) related to Hubble telescope, Challenger disaster, aircraft accidents, etc.
While many of the stories could be scary, overall, the author is trying to maintain a tone of levity. It is also heartening to note that institutions related to the field of engineering usually take these errors seriously and are trying to come up with solutions that will prevent disasters. The approach to addressing these issues is discussed from the Swiss Cheese model. This approach expects some amount of errors to creep in in each layer of implementation (e.g. design, installation, usage). If these errors are seen like holes in a slice of Swiss cheese, ensuring that when the slices are used together the holes do NOT line up can still prevent disasters. Thus, a design error can be blocked by better processes used in installation stage or the usage stage. Parker points out that unlike the field of engineering that tries to develop better methodologies at the system level, field such as medicine, still tries to blame individuals for errors/blunders, prevent them from working again, claiming the issue is resolved, without looking at systemic, holistic solutions. One can only hope this will change as the world matures.
Book has a few illustrations and tables, that were hard to see on the Kindle edition, which was a big disappointment. At times the authors attempted humor may fall flat. But overall, it is a good fun book, particularly engineering/medicine students should read. Will teach us to be humble.
No comments:
Post a Comment