I have been interested in the history of technology for a long time, and more specifically in what new technology does to the society that produces it. That is what led me to Labatut. He writes about scientists and engineers, but he does it the way an anthropologist would, which is not how these stories normally get told.

This was the first of his books I read, and I liked it enough that I picked up another one five weeks later.

What it is

The MANIAC is mostly about John von Neumann, told through the people around him rather than through him directly. It moves from Budapest to Göttingen to Los Alamos to Princeton, and then jumps forward to the Lee Sedol matches against AlphaGo at the end.

Labatut does deep research and then fictionalizes the interior parts, imagining how these people actually felt at the time. Some readers find that objectionable. I think it makes for a far more interesting book than a straight biography would have been, and I will come back to that at the end.

Highlights from the book

The book opens with Paul Ehrenfest, and there is an early description of how he worked that I keep thinking about. He was after what he called der springende Punkt, the leaping point, and deriving a result logically was never enough for him:

“That is like dancing on one leg,” he would say, “when the essence lies in recognizing connections, meanings and associations in every direction.”

That is a good description of the difference between being correct and actually understanding something, and I have met very few engineers who operate that way.

There is a passage about pure mathematics that surprised me, because it treats the subject as something closer to a religious practice than an academic one. The argument is that mathematics should be approached carefully, “because it has true power, a power that can be easily used for evil.” Given where the book is headed, that reads as a setup rather than a flourish.

The part I highlighted hardest is a small, bitter argument between the engineers and the physicists at Los Alamos:

‘Cause we actually soldered things, burned our hands, while they would just roam around like dinosaurs with their heads in the clouds trying to unravel the mysteries of the universe. We? We were building something. Something that would change their world. And they despised us for it.

I don’t think this tension has gone anywhere, it survives in our everyday struggle between engineers and researchers.

Then there is the line about the machine itself, which is the one I have quoted to other people most often since reading it:

“This species of device is so radically new that many of its uses will become clear only after it’s been put into operation.”

I don’t think that was meant as a warning at the time. It reads like one now.

A few other things I marked, more briefly:

  • Von Neumann came up with the Monte Carlo method while playing solitaire with his mind essentially idle. He noticed he could roughly predict the outcome of a game after only a few cards, and turned that into a way of solving problems by random approximation rather than by working them out. I love that the origin story is boredom.
  • The bomb and the computer accelerated each other. The race to build the weapon was sped up by von Neumann wanting his machine, and the machine was sped up by the arms race. They are not two stories.
  • In a paper from the late 1940s he described a self-replicating “automaton” in three parts, including a decoder that reads instructions and builds the next copy. He was essentially describing how DNA and RNA work, years before anyone had seen the double helix.
  • There is a quieter thread about what replaced faith. Von Neumann apparently felt the loss directly: “We have no guiding star, nothing to look up or aspire to, so we are devolving, falling back into animality.”

Labatut is also unsparing about the mood of the people building all this. One line stayed with me: “All that maddening, unthinking optimism, all that cheerful naïveté under which they hid their cruelty.” I have heard versions of that optimism in our industry fairly recently.

Labatut vs. Biographers

The last third of the book is about AlphaGo, and Labatut tells it from Lee Sedol’s side rather than DeepMind’s. It’s thrilling, and it is thrilling precisely because he chose the person losing.

I read Sebastian Mallaby’s The Infinity Machine later in the year, which covers overlapping ground through Demis Hassabis. Hassabis is a genuinely fascinating person and the book somehow turns him into a complete bore. It’s dull and one-sided in a way that made me appreciate Labatut more.

That comparison is why I don’t mind the invented interiors. A well-researched biography can leave you knowing more and feeling nothing. Labatut’s method is less reliable and much more alive, and for this subject I think that is the right trade.

All quotes are from The MANIAC by Benjamín Labatut.