Member Reviews

As a descendant of a nuclear test veteran, the second I saw the cover of this one, I knew I had to read it.

This did feel to me that it was written like it was fact rather than fiction, and I know that there are blurred lines with what is fact and fiction here, but I wasn't sure about this.

The subject matters are eerie, and dark, but important.

Thought provoking, if not a little stressful!

Was this review helpful?

Labatut has written a book which is utterly compulsive... and wholly terrifying. His theme is the development of scientific learning, both the incomparable glory of human inventiveness and pure intellectual brilliance, and the horrifying spectre of what happens when that search for knowledge has no moral or humanistic bounds or limitations.

The ultimate journey is towards AI but there's already a gap between where the book ends and our world, ending as it does with self-learning Go/chess computers. The itinerary takes some time to get there though with much time spent with Janos/John von Neumann, mathematician, physicist and computer scientist. His contributions to twentieth-century knowledge are jaw-dropping: from pure mathematics and quantum mechanics to the development of the atom and hydrogen bombs, from game theory and the mathematical basis of rational decision-making as harnessed during the Cold War to the development of non-human/digital intelligence.

But what is chilling is von Neumann's eradication of all moral limits: he stands against other international nuclear physicists who sign a joint letter to Eisenhower saying that the atom bomb should never be used; he is the one who works out precisely how high above Hiroshima and Nagasaki the bombs should be detonated to cause maximum destruction and death; and he came up with the acronym MAD (Mutual Assured Destruction) during the Cold War with advice to launch a pre-emptive strike on the USSR in line with his game theory model.

The story is told via narratives from von Neumann's colleagues, friends, wives, and intellectual enemies. I wish there had been an afterword detailing what is fact and what fiction: Labatut notes his sources but otherwise doesn't comment on this point.

For me, though, the book changes once von Neumann dies and the focus turns to the development of AI gaming computers - I struggled to maintain much interest in the detailed game-by-game record of Go and chess against international masters. Up until this point, about 70-75% of the way through, this would have been a 5-star book for me but, somehow, despite the importance of the topic, the final quarter fell away.

Still, for most of the time this is a veritable page-turner, albeit one which I found stressful to read. It feels more wholly 'factual' than When We Cease to Understand the World and reveals those scary scenarios without comment.

Was this review helpful?