Ossus Library Index
Science Fiction Index


A novel by Roger MacBride-Allen
(1996, ACE Science Fiction)

Isaac Asimov's Robots, book 3

A plan to revitalize the planet's ecosystem by dropping a comet to the surface threatens to destabilize all of Inferno's robots.


-- First reading (hardcover)
January 16th to 23rd, 2001


I was somewhat neutral on this one when I started writing this review.  As it progressed, I liked the book less and less. While it had its good moments, it was rather dull, showing that robot stories are best when they deal with murder.

Spoiler review:

I don't really like Kresh as governor of Inferno. I much prefer him as sheriff. Devray, the new sheriff, does a good job, but not as good as Kresh did. 

The story was rather complicated, though it can be summarized quite simply. A scientist named Lentrall has discovered a comet, which, if broken up in to small enough parts, could be diverted to impact the planet and dig a canal from the southern ocean to the polar depression, which will likely fix the planet's ecological woes once and for all. 

This leads to the Settler population's attempt to kidnap him, which was the best part of the book. The setup was air-tight from the Settler point of view. They've seen the results of crashing a comet onto a planet, and they don't want to see Inferno come to the same fate. For some strange reason, though, they never talk to Kresh about it -never tell him about their experience with these doomed ideas. The sheriff figures out what is happening right away, sends a message to Lentrall's robot that the man is in danger, and the robot protects him, thus foiling the kidnapping. Another strange point is that every single Settler got away after the fiasco. Robots were running everywhere, grabbing humans and not letting go until they were safely out of the path of falling debris. Is it not likely that at least one would be caught?

But the Settlers were not finished. They erased all of Lentrall's files on the comet, as well as destroying his office. The man is so shaken that he doesn't even realize that his robot has a copy of much of the data. When he wakes up in the morning, though, he realizes that his robot should have volunteered the information. But it didn't, because it knows that crashing a comet onto the planet would doubtlessly harm humans. He calls Freda Leving, robot designer, but even she finds that questioning the robot is futile, and the robot dies. I truly enjoyed this part, too. It was reminiscent of Susan Calvin's robo-psychologist methods in the early robot stories.

Enter the head of security for the Ironheads, a political group that thinks robots should do every single thing for every person. He is vehemently against Freda's New Law robots, and has vowed to destroy Caliban, the robot with no laws. The new law robots have taken up residence in a secret underground city, Valhalla. The Ironheads know only that Valhalla is in Inferno's Utopia region. And that's where the comet would strike if Kresh decided to go ahead with the plan. What could be better? The Ironhead security chief had raided Lentrall's office earlier, and is thus able to provide Kresh with all of the missing information. 

Maybe the Settlers couldn't do the job of crashing a comet onto the planet by themselves, but with a robot's help, the Spacers think it can be done. And this is where the robots finally come in. It took a long time to get to the robots, and that's the main problem with this book. The robots are incidentals. It was as if the author set up such a large problem in the first two books that he needed closure for it. Unfortunately, I think he provided closure for the other parts of the story, and less so for this part. The New Law robots could have been a major boon or pain. As it was, only Prospero got any attention. 

For Prospero, leader of the New Law robots, is crazy, and he doesn't want the comet to destroy his home. In the day just before the comet strikes, he kidnaps the leader of the Ironheads, and is ready to have him die. The robot cannot kill the human directly, but through his insane interpretation of the First New Law, he could set up the traps so that he could watch the human kill himself by setting off the traps trying to escape. Luckily Caliban figures this out, and kills Prospero, rescuing the human just minutes before the comet strikes. 

The main analysis of robotic behaviour in this book comes from the way they deal with the comet crash. There is Donald, assistant to Kresh, who has no problem with the comet crash until the Ironhead leader is kidnapped. Now he has a focus point, and he goes out of control in trying to send other robots to save the man, and in trying to stop the comet. There's Prospero and Caliban, who don't really have much to do until the very end, where the situation pops up out of nowhere. There's Lentrall's robot, who shows that it's dangerous to rely so heavily on robots. There are the hundreds of robots who have been dealing with the crisis for so long that when they hear of the kidnapping, they are suddenly relieved to have a focus, like Donald, that they go running into the middle of nowhere in a blind search. 

Finally, there's Unit Dee, who together with the non-robotic Unit Dum, controls the reterraforming effort. She has been convinced that the entire thing is a simulation, so that there is no danger to humans. But once Donald tells her the truth, she seizes up. She gets out of the conflict when she realizes that she has to make a choice about the least amount of damage, but for that she has to rely on the judgment of Kresh. 

What I don't like here is the implication that simply by telling a robot that a person is in danger makes the robot conflicted. It seems too extreme, even in a Spacer society. Donald tells the world that a human has been kidnapped, and the robots all go out of control. This is a robot talking. Shouldn't they try and verify the information first? I thought the same about Unit Dee, except that she had expressed suspicions earlier, so Donald had just given her a new input.

Ultimately, the book did not focus enough on the robots. There was some sloppy closure at the end for the New Law robots, and presumably Caliban, when Freda realizes they were a mistake, and that she won't let any more be built, and they should not be repaired when they break down. In the last book, they tried to put some restrictions on the New Law robots, but humans decided to smuggle them out anyway. The same thing will happen again, without a doubt. And if a few escape the planet, they could start a new cycle of what happened on Inferno. How would anybody trust robots again?

The scientist also gets some sloppy closure, and I can't figure out why. He goes crazy. From the insolent and superior-minded person that he was at the beginning, he crashes when his robot dies, and he decides later that he was wrong, and goes quite mad -ending up putting posters on walls that say "stop the comet!" And that's the last we hear of him. I guess the author couldn't figure out what to do with him after the kidnapping attempt. 

The book picked up its pace more than halfway through. Except for the attempted kidnapping, the beginning never got off the ground. I was ready to skip whole chapters when Caliban and Prospero were around. Their situation never appealed to me, and they only served to set up the fact that Prospero was crazy. Caliban, for all the exposure he had in the first book, barely existed until the last few pages. 

In all, I can't justify giving this book a passing mark. It had its exciting parts, but they were too little, too late. I enjoyed all of the parts where the police were involved, but they were a minority. It seems to imply that robot stories are best framed around murder, as we try to decide whether a Three Law robot is capable of committing the act.


Back to Top

All reviews and page designs at this site Copyright © 1999 -  by Warren Dunn, all rights reserved.