author : Donna Andrews

Turing Hopper is worried. Zack hasn't been at work for several days, and no-one seems to know where he is. She needs to investigate further, but she has a problem: she's an AI confined within the company's mainframe. So she recruits the only two humans who know she's sentient to help her: Maude the fifty-something PA, and Tim the Xeroxer and wannabe gumshoe. Soon it becomes clear there is more at stake than simply finding Zack, and Turing realises the lives of her friends, and even her own existence, may be in jeopardy.

On the surface this is a fairly conventional lightweight detective thriller, leaning more towards the style of Mission: Impossible than of Miss Marple. There's lots of excitement and chasing around, trying to outwit corporate goons. But what saves it from being pure mind-candy is the protagonist: the engaging character of Turing the AIP. Most of the story unfolds from her internal perspective, as she develops beyond pure sentience, to an entity fully engaged with the real world. After willingly suspending disbelief about the initial onset of sentience, I enjoyed the neat technical uses of memory, modems, email, databases, voice synthesis, and waldoes. We also get wry asides on all the things a real Turing test should cover, some bizarre recipes, what scares different entities, and questions about the nature of identity and the reality of sentience. I hope Andrews hasn't used up all her ideas in this first book, and can keep up this character development in future tales. I'm off to get the next one to see!

In Turing Hopper's second excursion, the now fully sentient AIP has a problem when one of her staff turns up dead. [As ever, don't work for, be a neighbour of, fall in love with, or be related to, any fictional detective.] When the police brush it off as a drug-related mugging, she, Maude and Tim are left to discover what really happened. They are quickly plunged into a world of secret identities and sinister role playing games.

More detection and less internal AI ruminations than the first book. The plot is unnecessarily complex, with at least three separate peculiarities going on, not really allowing sufficient clues to be planted for any one of them. So the identity of one of the bad guys is deducible by the usual "who is present in the story for no readily discernible reason" heuristic. I like the emphasis that it's the people, not the technology, responsible for the various crimes, but I miss the more philosophical thoughts about the consequences of sentient AI. However, on this note, there is a wonderfully grim set-up for the next book.

Turing Hopper, sentient AIP, is searching for her kidnapped clone/twin, T2. When the credit card of the main baddie is used, she thinks she might have a lead. But it leads her, Maude and Tim into a world of credit card fraud that ends in murder.

This is fun detective story, with lots of AI musings, and amusing gardening interludes. The various viewpoints show how Turing and Maude are completely misunderstanding and misinterpreting each other, and Turing comes over as semi-autistic at times, though desperately attempting to understand her human friends. The fraud case here is solved, but the wider problem of T2 remains, due to the somewhat Hamlet-esque ending.

Turing Hopper, Artificial Intelligence and Private Eye, is back with her crew, investigating a possible computer fraud, uncovered because of a hit and run. But the deeper they look, the more confusing things become. And there are hints that her clone/twin T2 might be sending a covert cry for help.

More fun detecting, with a high load of computer jargon, as we delve into the world of web-hosting, spam emails, and cats. There's a lot less chasing around and physical danger this time, as most of the action takes place in cyberspace, or in lecture mode. But there is a definitely meat-space ending.