Robert J. Sawyer is a great science fiction writer, having won every major award in the US, UK, Canada, Japan, and would have won one in Antartica if they had a contest. This novel won the Nebula and was a Finalist for the Hugo.
Frankly, I do not see why.
The story is based on two scientific premises: detection of the soul leaving the body and computer based artificial intelligence. Detection of the soul leads to experiments in AI to determine what life after death might be like. Dr. Peter Hobson, the inventor of the "soulwave" detection, uses AI and nueral net scanning to create three versions of himself: a life after death sim, an immortality sim and a control sim that is just like him. Hobson has some issues to deal with in his personal life (I won't play spoiler here), and those issues are duplicated into the three sims. One of them goes bad, and starts using the net to kill people.
Sawyer's claim to fame is that he will take premises like this and wrap very real characters around them. The concept of science fiction is in making both the science and the fiction work for the reader. Many writers tend to forget this, either throwing out unbelievable science or getting the science right but forsaking the characters or the plot. Sawyer is normally magic in this.
The Terminal Experiment is a good read, with nice pacing. It bogs down at times in the explanations of the science, and some of the philisophical discussions of the AI's. But the concept of killer AI computers has been hashed and re-hashed (remember HAL!), as has the concept of detecting something that proved life after death. And unlike other Sawyer novels, I had difficulty caring about the characters, esp. Cathy, Peter's wife.
I'm glad I read it, but I'm gonna go now and read Hominids, Humans and Hybrids, his classic Neaderthal Parallax series.