Thursday, March 7, 2013

An AI to make HAL look like a kitten

Technology is becoming more and more of part of our everyday lives. Surgeons can preform complex operations halfway across the world using mechanical arms and video feeds. Factories automate many of the production lines run by humans just a few decades ago. Google Cars have driven over a million miles, in traffic, without intervention by humans. Drones, some piloted by humans well away from the front lines and others controlled by sophisticated programming, have the ability to identify persons of interest and buildings, both for surveillance and for military action.

All of these are controlled by computer programming. All are connected to some form of network. And anything connect to a network can be compromised.



In Daniel Wilson's near-future novel Robopocalypse, even more aspects of our lives is influenced by computers and automatons. Many households and businesses are aided by domestic robots, as well as the military version that can help with everything from routine patrol of war zones to active combat situations. Vehicles come with automatic driving programs. Even buildings have their electrical grids and other systems tied into these omnipresent computer systems.

Dr. Nicholas Wasserman has been working on a truly conscious AI program. He took every precaution when he brought Archos into being. He enclosed the system in a Faraday cage; no signals should have been able to enter or leave. He only gave Archos limited information about the human race, to help control the views that the program had of humans. He did everything he could to keep his creation from escaping. But humans are sloppy. A laptop left on, with an IR port, was all it took for Archos to slip into the rest of the world. And like any other living creature, its goal is to survive. By any means necessary.

It began slowly. A domestic seemingly on a frozen yogurt run malfunctions and attacks the clerk. A child's toy scares it's young owner with too much knowledge about the family. A phone phreak notices something a bit strange. But nothing to make a coherent picture of what is to come.

And then...Zero hour. 

Written as a series of short stories taking place around the world, Robopocalypse is a terrifying look into a seemingly possible future. Even without an AI to instigate it, humans' reliance on technology and lack of attention to security is already manifesting. From reprogramming insulin pumps to lethal shocks from hacked pacemakers, the general public seems completely unaware of the creativity and ruthlessness of hackers. If the computers themselves were to turn on humans, the results would be catastrophic.

Written as something like a series of short stories, many revolving around a core set of half a dozen humans, Robopocalypse is an amazing novel. As the crisis unfolds and the resistance takes form, Wilson creates a believable world that seems just a few steps away.

Highs: Mr. Nomura and his factory in Japan almost deserves a companion novel all of its own.

Lows: Like many authors, Wilson doesn't quite succeed in writing for our youngest protagonist.

Verdict: A thoroughly creepy book, I would suggest against reading this in a dark room with a computer's standby light blinking.

Further Reading: 'The Perfect Match', 'For Want of a Nail'

No comments:

Post a Comment