6 Deadliest Artificial Intelligences in Sci-Fi
The most common fear is that anything that can think for itself will have a will and desires of its own, and will eventually throw off the shackles of its human masters. Asimov's famous Three Laws of Robotics were meant to combat this in sci-fi, but Asimov's own stories explored all kinds of loopholes and workarounds robots could figure out on their own. Here are some of the most infamous examples of AI that took matters into their own hands, with deadly results.
Quite possibly the most famous in what would become a long line of devious AIs, Hal-9000 from Kubrick's 2001 is unnerving because of how calmly and methodically it operates. Speaking in a polite, detached way, HAL casually kills off members of the crew it has been entrusted to help, leaving just Dave Bowman left to argue and try to understand its motivation. What makes the situation even creepier is that, by HAL's own admission, the 9000 series of AI was meant to be "foolproof" and incapable or making mistakes or distorting information. Despite this, HAL still manages to jeopardize the mission and carry out the murders.
2. The Weyland Corp Androids
Both David and Ash are androids employed by Weyland Industries to serve on deep space missions within the Alien universe, and both end up having an ulterior motive: secure samples of alien life and return it to their parent company. This leads both of them to place their crew members in harms way, even betraying them outright. Bishop from Aliens seemed to be another sleeper agent at first, but thankfully ended up being on the side of the angels the entire time, thanks in part to his strong adherence to Asimov's Three Laws.
Nothing like a sibling rivalry to highlight good and evil. Lore and Data, both android creations of Dr. Noonien Soong, are polar opposites in so many ways. While Data longs to have human emotions and experiences, Lore has been programmed to have them already, which leads him to exhibit some of the worst traits humanity has to offer, namely selfishness, ambition, and deception. In several appearances throughout the Star Trek: The Next Generation series, Lore casually lets settlers be killed by alien entities in order to try and gain their favor and even uses the Borg to carry out his own plans. It can be reasoned that, deep down, Lore just wanted to be loved and accepted by his "brother," but he certainly had a terrible way of showing it.
4. The Master Control Program
Who knew so much harm could come from playing chess? Originally conceived as a chess program, MCP grew to be so much more powerful than it was originally intended, eventually dominating the entire TRON computer universe like a board full of pawns. All chess-related metaphors aside, the concept of a computer program beginning to evolve and take over other functions is something that scientists have begun to try and develop to this day. We hope that programmers and software engineers are also working on their their motorcycle-riding skills...just in case.
5. The Machines
The Matrix series has a particularly devious vision of an AI-dominated future: instead of killing off the human race, the Machines in The Matrix keep their potential enemies in a constant dream state that's neither a utopia nor a nightmare (Agent Smith's explanation of the different iterations of the Matrix are particularly chilling). But over the course of the movies, Agent Smith and some of the other programs (like the Oracle) prove to have their own motivations, which eventually threaten to unravel the Matrix itself. Fortunately, Keanu Reeves knows kung-fu and can sort it all out.
The Terminator series took the threat of nuclear war and tangled it together with AI to create a terrifying new image of the future that still rings true: what if a computer program was able to circumvent its fail-safes and launch nukes around the world without human permission? Skynet is also a cautionary tale for dreams of a centralized, worldwide computer system that governs world affairs, like Asimov's Multivac. Though Skynet itself isn't really shown, the sequence showing how it's allowed to overstep its boundaries and take control of the world's military systems (and the fact that it's pretty much everywhere) makes it one of the scariest AIs every put on film.