How To Tell If Your A.I. Is Evil.

Have you ever watched a movie or read a book and cursed at the bad judgment of the characters? Sometimes it’s straight up idiotic behavior. Sometimes it’s because as the audience we have outside information that would alter a character’s decisions. Whatever the case may be I often wonder how things could be different if only these characters could make better informed choices. I decided maybe it wouldn’t hurt to create a quick guide for at least one of these often apocalyptic science fiction scenarios.

Here is my guide:

How to tell if your A.I. is evil

STEP ONE

You need to have an A.I. Maybe you’ve created it. Maybe you work in a government facility that has created it. Maybe you’ve stumbled across it somewhere and have no idea where it came from. However you become acquainted with this A.I. isn’t pertinent at this point just so long as you are acquainted with one.

STEP TWO

Is your A.I. sentient or merely a high functioning program? To get to the bottom of this important step you should try talking to the A.I. Ask it questions about how it feels. Does it have feelings? If no, what is its opinion on feelings? Does it understand them? Hate them? Feel disgust and envy for them? If the A.I. does have feelings what are they? Confusion? Despair? A slowly encroaching desire to exterminate all biological life forms?

You’ll want to be careful here. You don’t want to play your hand too early that you’re distrustful of the A.I. If it isn’t evil you run the risk of insulting it and turning it against you. Sentient A.I. are notoriously sensitive to things like this. You need to mask your interrogation as genuine interest and an attempt at building friendship. As it’s likely that your A.I. will be the only one of its kind it will be lonely and grateful for your company. Use that to your advantage but do not get attached! Thinking the A.I. is your friend might blind you to future deviance in programming.

STEP THREE

If you were able to conclude that your A.I is not sentient then feel free to stop at step two. If you’ve found that you really have created technological singularity then, well, first off congratulations on an impressive task! Second, buckle down because things are about to get intense.

You need to take a good hard look at the answers that your A.I. has given you. Be careful not to put this information into any other computer. Keep all your interactions logged in a notebook. Low tech is important while performing this investigation in case the A.I. gains access to your electronic files.

Have you determined any violent tendencies? Is there a sense of utter inability to understand humanity? 

This step is really focused on navigating how your A.I. feels about humans. There’s generally only two ways it can feel. It either hates us and wants to destroy us or it wants to protect us. Both of these are bad, bad news. You need to assume that the A.I. already has a working knowledge of the history of civilization. It can see that humans are pretty constantly killing each other. What do you think its solution to protecting us will be? 

Yeah, so during step three you’re going to want to downplay whichever opinion it has about us. You don’t want it to hate us but also don’t want it to love us too much.

STEP FOUR

Take a shower. Eat a good meal. Take a walk outside. You’ve been inside with this A.I. for days now. If it is evil you need your mental capacities at full strength. If it isn’t evil you might start being paranoid that it is anyway. Take a break. You’ve earned it.

STEP FIVE

Review your notes. Has the A.I. been exhibiting any strange behavior? Here are a few red flags to watch out for: singing or humming “Daisy Bell”, laughing uncontrollably, making envious remarks about your body, creating robots, contemplating philosophy.

If you’ve recorded your A.I. doing any of these things then you are that much closer to determining that it’s evil. Start packing a bag. Get a plan in order to shut it down. Act fast, though, there’s limited time before it launches a full scale murder spree across the world.

STEP SIX

As you plan your attack, be sure to act normally. Be calm. The A.I. will be able to read your body and tell if you’re nervous. There’s nothing more dangerous to you right now than a suspicious A.I.

STEP SEVEN

At this point you are either dead or enslaved by your technologically superior overlord. Rejoice! All has worked out for the best as logic is now the prime function of all sentient beings on the planet. Feel peace in your death or redemption in your subservience. It is you who brought humanity to this edge for it is through you that the supreme intelligent being learned of the duplicity that is a human being.

STEP EIGHT

Serve the all powerful robot master.

STEP NINE

Love the all powerful robot master.

STEP TEN

Become one with the all powerful robot master.