Tips for Writing a Novel with Artificial Intelligence as a Villain
From One Author to Another
In many sci-fi movies and novels, the common theme of a man-made horror of the web is often incorporated as the driving factor for the plot's progression as the indifferent super-intelligent life-form turns on its creators, unhindered by the concepts of morals or attachment that the protagonist fights to protect. However, as the theme has grown increasingly more common, fans of the genre begin to raise their standards for the authenticity of A.I, looking for more than a simple terror of the night as the same old story begins to grow stale.
The story-writer's main issue with this scrutiny comes from the very thing that makes A.I so scary; they don't make mistakes. Unlike humans who can miss details, or make great leaps in logic from pumped up emotions or just poor planning, our coded menace thinks far too quickly to have any excuse for errors, and lacks the human emotions that would blind it to its own idiocy. As a result, the simple tricks that many writers employ to give the protagonist a fighting chance are unacceptable. A weak point in the death star or an arrogant villain who likes to talk himself for slightly too long just don't fit in for the perfectionism embedded in the mind of a computer. What's more, the complexity required to produce artificial intelligence in the first place means that a glitch or fatal flaw cannot be wafted in the creator's direction either. Even a safety protocol falls short against A.I. who can easily scan their own code to correct their "imperfections." So how do you create a being so unbeatable and still allow it to lose, or at least appear to stand a chance of losing?
As it turns out, there are a few methods of beating this problem. However, unlike the easy methods from other genres, these ones require tackling the problem head on, with no Deus Ex Machinas to save the day.
The first solution is to have the A.I. win. Sounds boring, but it can be presented as a twist, where the protagonist thinks they are winning until they fall into a trap, or find they've been working for the wrong side the whole time. This is the easiest method as the A.I. is impossible to question in its authenticity as it didn't make any errors and its plan went as calculated. Examples where this has worked in our media include the classic 2001, A Space Odyssey and more recently Ex Machina. However, this method leaves little room for expansion on one plot and is already beginning to enter the realms of being a predictable plot. But don't lose hope yet. There's still time for plenty more exciting stories of this type and the narrow fit for the plot only challenges writers to explore new ways of creating a spin on the story to leave their own unique mark on the genre.
The second solution to the problem is to have the AI not be interested or have no quarrel with the protagonist. Now, this doesn't mean that the A.I. will act as a side character, but it does mean that there has to be another reason for the A.I. to be in the protagonist's way. This would be in the form of protocols or firewalls that stop the main character, or the A.I. possessing items or abilities required by the protagonist that the A.I. is not ready to give. Examples of this include the matrix series character "Mr Anderson," who has no problem with many of the protagonist's, Neo, actions once he attains independence, but reserves a separate interest in controlling the matrix, which is not Neo's concern. However, as both groups attempt their own goals, they regularly collide until they are forced to confront each other and defeat the other in order to reach their goal. The difference here is that Mr. Anderson slowly begins to act less like A.I as he gains independence and loses because of an event he could not predict rather than simply not having the preparations to deal with an unexplained attacker. Unfortunately, this method fails to produce much conflict as the protagonist can generally wait until they see a weak point and quickly assure victory which is why there are not many popular stories that use this method. However, this way can lead to secondary story arcs to produce a more complex story as their time lax narratives allow for sidetracking and multitasking.
The final solution is to humanize the A.I., which is the most common approach. By having the A.I. learn from, or be based off of the human mind, the presence of emotions becomes more acceptable, which in turn allows mistakes to be made. Now these mistakes are limited due to the A.I.'s continuing ability to process information at super speed, but bouts of anger or attachment can lead to risk taking or inefficient solutions that give the protagonists a fighting chance. There are a multitude of examples of this method being used, but some particularly effective instances among them include Marvel's Avengers, Age Of Ultron, the meta from Red Vs Blue and multiple instances of Borg-human relations in Star Trek.
Please leave your own opinions in the comments and feel free to take a look at my own novel on Amazon kindle E-books in the link below. Feedback on future topics to cover is also appreciated.