I thought it was a good book and helped further explain the movie. It added more detail to the scenes. Its action packed and hard to put down. This novel will be a favorite for any terminator or schwarzenegger fan. While the movie may not be as good as the others, the book is just as good as any.

If you have not seen the movie, I would highly recommend the book. It describes many of the scenes in great detail. Also, some events in the movie that are a bit unclear are explained quite well in the book. For instance, when the T-X reprograms Terminator's memory system, it would seem as though it would be impossible for the Terminator to be on John Conner's side again. But the book explains that the Terminator re-booted his computer system, and thus was able to have a fresh start. In the movie, this is not explained at all, and the Terminator just comes back to save John Conner, which appears puzzling since he was, at that point in time, programmed to harm John Conner.

The book is very action-focused - with very vivid descriptions of the actions that are occuring. I am actually quite impressed with the ability to write a book based on a movie of this complexity -- and still make it very readable, exciting, and a fun read.

-- Michael Gordon
Los Angeles

Writers though have expressed considerable enthusiasm regarding AI, and this book, and the movie that accompanies it, is ample proof. If only the field was advanced as this book portrays it to be. Concrete results and applications of AI though are currently accelerating, and there is little doubt that battlefield robots will be a natural consequence of the current AI technology.

The book illuminates to some extent the method of time travel that was not discussed in the movie: the Hawking/Einstein wormhole scenario but generalized to superstrings. The superstring wormhole/time travel machine was discovered in the story by a graduate school at Oxford...an incredible achievement for one individual, and even more astounding given the fact that current superstring theory has no experimental ramifications, except for predicting a huge value for the cosmological constant. To go from the current state of superstring theory to one where one can do spacetime engineering as a consequence is quite a leap in knowledge. The wormhole is opened by the focusing of sunlight using of all things a solar sail, which results in several hundred terawatts of energy over nanosecond time scales to arrive at the place of the singularity equipment. Objects are able to travel backward in time, and the time machine has a replica under human control.

The story has some plausibility in light of the current use of artificial intelligence in network engineering, especially network security, network event correlation, and network capacity planning. Indeed, it was announced this week that a technology is now available that will identify security risks and take action using auto-adapting artificial intelligence. The story makes Skynet one of these smart network applications, so intelligent in fact that it becomes "self-aware", gets paranoid about human intentions, and therefore orders a massive nuclear strike in order to remove the human threat. This move by Skynet makes the story somewhat implausible, for if, as the story holds, there is no "central core" to Skynet, it being instead a distributed application that runs on computers all over the world, then it would destroy itself in the very act of a global nuclear strike. It would have been better for Skynet to "lay low" and make sure power systems cannot be tampered with instead of ordering such a self-destrucutive act. It is the power systems that are most crucial for the survival of Skynet, and its distributed nature requires such power sources to be left intact globally, and not just "under the mountain" where its inventors program it. In addition, there is no need in the story for Skynet to become "self-aware" in order for it to engage in reasoning that will protect it from harm. The agents and spiders it moves around in the global Internet could make logical deductions to this effect. Such agents would then spend most of their time insuring that power supplies are redundant enough to keep Skynet's global nature intact.

The action in the story is typical of the Terminator movies and book series, with the female-emulating TX Terminator robot, highly sophisticated technologically, taking the story for sure in this regard. But the story also captures the introspection of John Connor, the main character and hero, and the one responsible for leading the future war against the machines. A human being facing this knowledge of the future would be under considerable stress, and this is brought out in the story via his dreams. The dreams are of a nightmarish future, with a devastating war of humans against machines, a war that Connor and his lieutenants will eventually win, much to the chagrin of the machines. The machines can't accept their defeat, and consequently send replicas of themselves through time to try and kill Connor and his lieutenants.

Should we label the machines as intelligent considering their behavior? Do intelligent entities engage in the violence and horror that these machines do? One can of course imagine schemes and plans that might justify such behavior, but a more practical strategy would be to ignore human interactions, or possibly engage in a mutual symbiosis. Intelligent entities realize the waste of resources and intellect in the making of violent confrontation, using it only as last resort. There are so many scenarios that would be more optimal for the course of action of these machines, and it would not be a credible argument to hold that they act as they do because of their training via humans, considering the relative sparsity of human violence throughout history. One should interpret therefore the machine decision for war as a mistake, and not one that is practical, and therefore not moral. They failed to seek alternatives that would insure their survival, and this is ample proof that they are not intelligent, or at best marginally so.

The book though in a sense is a portent, however inaccurate, of things to come, and things that are happening right now in artificial intelligence. We do not have robot armies, but we have AI invading many domains: financial engineering, network engineering, mathematics, physics, Ecommerce, bioinformatics, to name just a few. The applications of AI are accelerating, and there is every indication that this trend will continue. We are entering a world of the silicon geniuses, the world of the avatars: we are indeed witnessing, and are priveleged to do so, the rise of the machines...