Review: Foucault's Pendulum

Foucault's Pendulum by Umberto Eco
My rating: 5 of 5 stars

This book changed my life. When I read it, I was playing around with a neural net, a kind of logical "black box" that can find patterns. Of course I was trying to predict the price movement of stocks. After literally hundreds of minor adjustments to my model, I got a stock predictor that actually worked. I traded it based on the indicators once and made some quick money. But then a lesson from Foucault's Pendulum started nagging me, the talk about the laundry list. (view spoiler)[ In the book, a researcher finds an old piece of paper, full of holes due to age. He can make out writing, but many parts of words seem to be missing. The researcher works on deciphering what the paper was about, assuming it was related to the Knights Templar, and he comes to realize that, if those missing pieces are filled in with certain letters, the document is a key to everlasting life and a timeline to a Templar reunion. This belief becomes pivotal in the story, leading to the researchers eventual demise when others demand the proof. But just before this happens, the researcher's wife points out that if you replaced the missing letters with different ones than the researcher chose, you would see that this was a centuries-old laundry list. In other words, you can make many stories from incomplete information, and an incorrect interpretation could be deadly. Occam's razor, dude, Occam's razor. (hide spoiler)]I decided to stop working with the black box, although I think I could have used it further, and possibly made a lot of money trading or writing about it. I had started to believe the story the black box "proved". At least I quit while I was ahead! Although I kinda wished I hadn't... BTW, I was predicting ERTS using CPQ (the former Compaq) and DELL as proxies for the market over trades between one and ten days. Back in the mid 90s. On a 386 using one bootable diskette.

View all my reviews