Latest Club Soda cross-postings

I’ve just published two blogs at Club Soda. One is on the relationship between alcohol and cancer, and the other a background piece on epidemiological methodology.

And last week I wrote a little (tragi-) comedy piece about hunting for non-alcoholic beers at a beer festival.


Book notes: Social networked physics

Paul Ormerod: Positive Linking – How Networks are Revolutionising your World

Alex Pentland: Social Physics – How Good Ideas Spread

Two recent reads on similar themes. Paul Ormerod has written a few interesting books about economics and networks already (starting with The Death of Economics), and his latest from a couple of years ago continues on similar themes. Alex Pentland’s book from last year is also concerned with networks. Both books also follow a similar structure: a bunch of interesting studies, examples, and stories to begin with, followed by a more general discussion of “what ought to be done” at the end.

Where they differ is in their confidence. Ormerod talks a lot about the unpredictable nature of human behaviour, and is (rightly, I think) skeptical of any central government being able to manage things as complex as a national economy. Pentland, on the other hand, is convinced that with enough big data and some clever mining of it, any social problem is just waiting to be solved. Absolutely, the methods and findings he presents in the first half of his book are very interesting, but I couldn’t help thinking that he is more or less dismissing almost everything ever done in any of the social sciences (plus psychology and a few other fields as well). For example, he ignores more or less all of economics since Adam Smith: I’m no great fan of much of economic theory either, but there is still much insight there to be found. And it would be easy to pick holes in his logical jumps from a handful of studies to wide-reaching conclusions about human nature and behaviour. Plus all the big data critiques still apply.

But all in all, two stimulating books well worth reading.