Writing from the perspective a career Air Force officer and fighter pilot, Riza ably summarizes the literature on the laws of wars and autonomous systems (primarily Michael Ignatieff and Ron Arkin, but also Patrick Lin, Peter Singer, and many others). His twist is based on the existential credo of Sartre that "to kill one must be willing to die", and a warrior ethos that is vital both to the ability of officers to command and the legitimacy of winning wars against technologically inferior enemies. As Riza argues, the immunity that autonomous weapons platforms provides against harm may make wars easier to start and dialog leading to peace more difficult to develop. The overall thrust is that while autonomous weapons certainly *can* kill people, and may even be able to kill people in accordance with the laws of war, the use of autonomous lethal robots undermines the moral foundation which makes war and peace possible.

There are several moments in this book that are quite interesting. Riza's comments on the changes in air force manning and potential loss of tactical skill development in junior officers are a new take on the issue. However, I think that several major points are not addressed by this book.

First is the American orientation of < href=http://en.wikipedia.org/wiki/Battlespace>Battlespace Dominance. Since WW2, American military superiority has been based on the ability to bring overwhelming fire to bear on vital parts of the battlefield, while monitoring areas around strategic assets and allowing the whole system to manuever freely. Battlespace is a technocratic doctrine of making lists of target coordinates and knocking them down as efficiently as possible. How does this practice of war comport with warrior codes?

Second is a romanticization of air combat. Fighter pilots surely do have The Right Stuff and go in harm's way, but conversely air tactics are not about dueling but about clobbering the other guy as fast as possible, ideally getting a kill shot before they even know they're dead. And in the long occupation of Iraq and Afghanistan, American pilots faced essentially no opposition in the air.

And third is an under examination of robot vs robot combat. Al Qaeda doesn't have any robotic capability (unless you count IEDs), but a nation like China or Russia might. While losing an autonomous platform might not be the same as suffering human casualties, these systems aren't free, and are in fact quite expensive. I've heard that recent military exercises have involved trying to run down the enemy's inventory of precision guided munitions by offering bait targets at the edges of effective engagement envelops. Robot vs robot combat brings in the tricky issues of procurement policy and economic warfare, but it is substantially under theorized and could be addressed by someone with Riza's background.

I know it's a little unfair to criticize a book for what's not in, but I wanted a lot more. This is an area of emerging concern, and compared to Singer's Wired for War, Riza's book is less tied to specific programs and evolving robots, and more towards the big picture moral and legal issues in warfare.