The Information: A History, a Theory, a Flood

March 6, 2024 12:56 PM

A thoughtful and provocative exploration of the big ideas of the modern era: Information, communication, and information theory.

Reviews

Cory Doctorow:

The Information isn't just a natural history of a powerful idea; it embodies and transmits that idea, it is a vector for its memes (as Dawkins has it), and it is a toolkit for disassembling the world. It is a book that vibrates with excitement, and it transmits that excited vibration with very little signal loss. It is a wonder.

Ian Pindar, Guardian:

Too much information – information overload – is the subject of the last third of Gleick's book, where he quietly drops Shannon's concept of information as a mathematical abstraction independent of meaning for the every-day meaning of the word. This final section, the "flood" part, is something of a letdown because it lacks political awareness. Information overload requires us to sort the information we receive, says Gleick, to "filter and search", but filtering information, he lamely observes, raises issues "of trust and taste". We may well live in an "information age", but Gleick never asks who might be filtering the information before it reaches us. He regards "information poverty" as a problem solely for previous generations and not for viewers of, say, Fox News.

Awards (per Wikipedia)
2012 Royal Society Winton Prize for Science Books, winner.
2012 PEN/E. O. Wilson Literary Science Writing Award, winner.
2012 Andrew Carnegie Medal for Excellence in Nonfiction, finalist.
2012 Hessell-Tiltman Prize, winner.
2011 National Book Critics Circle Award, finalist (Nonfiction).
2011 Salon Book Award (Nonfiction).
2011 New York Times Bestseller
2011 Time Magazine's Best Books of the Year

posted by pwnguin (2 comments total)

I seem to have missed the initial buzz about this book, though I did happen to read Shannon's original paper A Mathematical Theory of Communication around the time this was published. Finally finished it last night after I saw this as a top recommendation on HN.

The first two sections do a remarkable job of laying the groundwork for the math by way of a history of human communication: African talking drums, alphabets, and dictionaries. Though he does claim at one point that all human language shares the same word for "alphabet" and I don't think that's quite true. Definitely not topics I would have researched on my own, but interesting nonetheless.

If you're unfamiliar with computer science theory, the middle section has a decent explanation, and if you are, Gleick offers a pretty good exploration of the personalities involved, and their circumstances. Kolmogorov, for example, appears to have wasted some of his time reproducing science the Soviets redacted from Shannon's paper in translation. What he doesn't mention is that his advisor was declared an enemy of the Soviet people by Stalin's regime, and may have had to testify against him.

But I have to agree with the Guardian review, on slightly different grounds: after exploring the tie between Shannon information and quantum physics, the "flood" section is less well tied together or fleshed out. He talks about filters but only briefly explains PageRank, and seems to completely neglect earlier, simpler filtering techniques like TF-IDF.

And of course, a decade on, the complaint quoted in a late section that "now we expect these machines to read for us" takes on a new salience. Especially since transformer based LLMs seem remarkably conceptually aligned with Shannon's Markov models from the 40's. Not that anyone could have seen it coming in 2011, but if there is a second edition some day (doubtful) this seems like the obvious extra chapter to add.
posted by pwnguin at 1:24 PM on March 6

I've had this on my TBR pile for a long time but still haven't gotten around to it. Reading this post, Pindar's review in the Guardian really strikes me as being a pretty damning critique to someone who wants to pick this up in 2024. But I'm interested enough in the history here that I'll probably still get around to it at some point.
posted by synecdoche at 4:23 PM on March 6