22 Dec 21

This book examines the development and use of lasers, electromagnetic pulses and cyber attacks in the context of the renewed great power rivalry and the arrival of partial artificial intelligence (AI). Del Monte is a US scientist and engineer who has led several classified projects and engaged with the US Department of Defense over many decades.

The parts of the book which can rely on the author’s expertise are excellent. He has the rare ability to explain complex physics clearly, both in themselves and in terms of the implications for some military and civilian affairs. He usefully combines this with shrewd insights from the inside of the US military-industrial complex. With the character of war continuing to change in a way that is difficult to fully comprehend, though perhaps not as fast as some breathlessly proclaim, this is a valuable contribution. There is also a useful summary of the US’s first to fourth offset strategies. These summarise the varying US searches for technological advantage since the Second World War: first nuclear weapons; second precision strike and stealth; third AI and autonomy; and fourth all of the above plus hypersonic weapons and the possible militarisation of space. The section on the Defense Advanced Research Projects Agency (DARPA) and its clones is also interesting.

The problems come when our author moves beyond his speciality. Del Monte is seemingly not aware that automated threat evaluation and weapon allocation was used in operations rooms long before the advent of the Aegis system, nor of the use of electronic sensory measures to detect radars decades before the 1980s. The Battle of Britain and the start of the Cold War are both misunderstood. Misnaming the current US lightweight torpedo is perhaps a detail, but sea mines are inexplicably missed from the discussion of autonomous weapons, when they offer a counter to some of what is said. There are other examples, and though none are absolutely central to the main point being made, they reduce one’s faith in the whole. A contributory factor to these problems may be the sources used: many of the footnotes lead to websites of variably reliability rather than peer reviewed publications. For example, almost none of the established literature on AI is cited. Also, in light of recent events in Afghanistan, it is slightly strange to read that US forces are the best at counterinsurgency in the world. This reflects a slightly narrow, US centric world view.

I don’t enjoy making this criticism for the author is clearly well intentioned. His stated hope that by writing the book he can make people aware of potential problems and that the world can then perhaps avoid a war using dangerous new technology should be taken at face value. He is a man in a hurry to avoid a possible disaster.

So War at the Speed of Light is a mixed bag. Much is excellent or at least a worthy contribution to the discussion. But the mistakes and the limitations in the research mean that one is constantly asking oneself “is this section reliable?” It’s an interesting book written with a positive intent, but it cannot be entirely recommended.