War in the Age of Intelligent Machines

Picture the following scenario: two countries, call them A and B have declared war on each other for reasons of their own (usually of course without moral justification). A begins the conflict by sending over to B's territory hundreds of thousands of entities that look like houseflies. Equipped with wings and tiny sensing devices, these entities swarm all over Country B and perform surveillance by communicating with each other and with A's central command. This is followed by a massive cyber attack on B, wherein its computing facilities and networks are flooded with intelligent network agents that work their way through B's networks and stymie their ability to route useful information. Once B's networks have been effectively made useless, A sends over thousands of ground-based and air-based robots, all of which are carrying ordinance and can identify enemy targets and destroy them as they see fit. These machines are able to communicate with other and coordinate attack plans, and they use both explosive power and cognitive disrupters to mentally confuse B's military infrastructure. At no time in this conflict was there any human in country A involved, except perhaps as a spectator. In fact, even the strategies and attack plans, as well as the decision to go to war against B in the first place, were the responsibility of intelligent machines. Country B, not having the same technological capabilities as A, is effectively decimated within a matter of days, if not hours.

This hypothetical and future-pointing scenario of the use of intelligent machines in warfare is one possible one that is not too far away from present capabilities. The author of this book gives his views on machine warfare, at least from the standpoint of what was available technologically at the time of publication of the book. It is an interesting book, possibly terrifying some readers, while inducing some to seek further information on just how effective machines would be in actual warfare. In this regard, this reviewer summarized the main points in the book as follows:

Machine Autonomy
Many times in the book the military is described as supporting the notion that the machines be given the authority to decide whether or not to destroy a target. Although there are no doubt many in the military that show support, there is still a great deal of resistance to this notion. Handing over to the machines the power to conduct a war as they see fit is still an idea that is hotly debated in the military and this is likely to continue unabated in the near future. The technology for conducting an autonomous machine war was not available at the time of publication of this book, but it is now. However the deployment of this type of technology, like many others, has been stymied by human anxieties and moralistic musings. But, as the author points out, a similar reluctance was shown historically in handing over control of battlefield events to the individual soldier. The author's use of language like "predatory computers" to describe autonomous war machines does not really help to calm the fears of those against the use of autonomous machines in real battle. But this use is an example of the radicalization of technology that this reviewer, along with a small minority of others supports without any mental reservation.

Chaos
Much space in the book is devoted to the "chaos" paradigm that was very popular at the time of publication. Originating in a branch of mathematics called dynamical systems, the characterization of a general dynamical system as being "chaotic" has made use of a wide variety of mathematical tools. Applications of chaos have found their way into statistical mechanics, economics, sociology, and cognitive neuroscience. The author evidently thinks the concept of chaos is important for his general case, due to his frequent discussions of it in the book. His main point, and one that this reviewer believes could be omitted without affecting his case, is that there is a connection between information-processing technology and self-organizing processes. This connection however is never really made clear in the book (and again could be omitted without damaging the author's main thesis).

Definition/Characterization of Machine Intelligence
The characterizing of a machine as intelligent has been problematic throughout the history of artificial intelligence, and this is also reflected in this book. The author usually attempts to characterize machine intelligence by illustrating instances where it is not present. Machines that merely perform logical deductions for example cannot produce new knowledge and therefore cannot be considered to be intelligent in the author's view. But machines that can execute inductive reasoning patterns are intelligent he asserts, since they can learn from experience.

Since the time of publication, inductive reasoning patterns have been realized in many different machines deployed in many different contexts and businesses. The author would refer to this as a successful "mechanization" of inductive logic, a task that he asserts would be very difficult to bring about. That is has is a testament to the ingenuity of the developers and researchers who brought it to fruition.

And this author, like others who write about artificial intelligence, cannot resist speculating about how to bring about more advanced levels of machine intelligence. As an example, he points to what he calls "information-based structures" such as DNA and enzymes, the workings of which he claims are similar to computer programs and information processing. His speculations have been borne out since the time of publication, with DNA and cell-membrane computing show to be compatible with the Turing computability paradigm and having enough computational power to solve hard problems. The author's discussion of the similarity between enzymes and program "demons" is now better expressed as a similarity between enzymes and agents. At the present time, the role of DNA computing in artificial intelligence is significant but it has yet not been applicable to bringing about general reasoning capabilities in intelligent machines.

History of Warfare
The author also devotes a significant portion of the book to the history of warfare, feeling that it is necessary to show how both regimentation and the relinquishing of command responsibility to the individual soldier have ramifications in the eventual acceptance of intelligent machines in actual combat. His discussions along these lines are interesting in that they point out the need for individual initiative on the battlefield as well as a centralized command-and-control. The machines will be expected to exercise initiative and creativity in the theater of war, and those abilities at the present time are not available. Not yet. But they will be soon.

How would a robot write a history of its own kind? This question kick starts De Landa's most famous book, detailing the history of the "war-machine" through the eyes of Deleuzian philosophy. I can't say I totally get Deleuze, but his project excites me in those rare moments I can parse it. De Landa does a great job of bringing Deleuze's philosophy into the real world and showing us how many of Deleuze's concepts can be made to do work in the disciplines of history and the social sciences. I can't really make any claims about how important this book is, or even how important De Landa and Deleuze are, but I love this book, it makes philosophy fun, kind of like watching a really great action movie, where ideas are the heroes. Happy to have it on my shelf.

great!

De Landa's theoretical connections are interesting and thought provoking but are unfortunately very often not adequately cited. In his attempt to explain the entire development of warfare, De Landa displays only a cursory knowledge of his topic frequently neglecting to mention crucial information. A few errors and mis-explanations early in the book made me question everything he had to say. This book is great for the images and connections it invokes. The theoretical leaps it makes are admirable. Do not however, rely upon this book for factual information.

Outdated and boring book.

Manuel DeLanda's preeminent virtue as a scholar is the way in which he applies the ideas of complexity theory (emergence, feedback, etc.) to the historical record, and this book follows this template, looking at moments where technological developments (the conoidal bullet, wireless technology) spur military systems to evolve (a process which, in turn, triggers other armies to evolve in response).

If you accept this premise (fail to at your peril), it naturally suggests that the militaries of today will one day evolve even further. So in addition to sketching out historical instances of this sort of thing, DeLanda spends a lot of time drawing attention to contemporary developments in technology or military theory that might be putting us on the road to future phase shifts that might spell Bad News for soldiers and civilians alike. Artificial intelligence, RAND-style war game simulators, and predatory machines (of the sort outlined in DARPA's "Strategic Computing Initiative") all come in for an extended critique, although DeLanda seems more optimistic about technological systems that don't take human beings "out of the loop" (the book ends with an appreciation of humanist interface designer Doug Engelbart).

All in all, this book is pretty essential reading for anyone interested in the "machine" part of the war machine, although it could definitely benefit from a little revision and expansion: some of the Cold War anxiety undergirding the book has lost some of its edge in the intervening years, and I could stand to lose some of it in favor of having DeLanda as a guide through past two wars (although War was published in 1991, Desert Storm hardly ranks a mention, a little odd, given the use of Israeli-built Pioneer UAVs in that conflict).