Hollywood Always Gets A.I. Wrong, And This is Why (An Analysis)

Warning: this post contains some spoilers for the movie Her.


From Metropolis to Terminator it is clear that Hollywood's crush with A.I. is not something new and this past year shows it is not ending anytime soon. Movies like Chappie, Ex Machina, Her and of course Age of Ultron help bring A.I. into the public consciousness, even big stinkers like Transcendence do the job of getting people to talk about the subject.

You probably have heard it on the news lately, big names like Bill Gates, Elon Musk and Stephen Hawking telling everyone we should be cautious and scared about A.I. This is happening because A.I. is clearly becoming more science than fiction, and this is not happening slowly.

A.I. is now on the news and on pop culture and since pop culture holds a tight grasp on how the public perceives science, this is the perfect time to explore just how it is that Hollywood depicts A.I., why this is a wrong depiction of A.I. and the motives behind Hollywood's misrepresentation of A.I.

How Hollywood Represents A.I. ?

Hollywood will generally depict two kind of A.I.:

1) Artificial General Intelligence: This is an Artificial Intelligence system that learns from experience, thinks abstractly, learns quickly and solves problems just like we humans do.

Chappie is a great example of a movie with an AGI as a protagonist, while yes Chappie manages to solve the problem of consciousness that is just a Deus Ex Machina for the movie, the moment Chappie achieves general human level intelligence he stops becoming more and more intelligent, and that is exactly the problem.

2) Artificial Super Intelligence: on the words of Nick Bostrom "This is an intellect that is much smarter than the best human brains in practically every field, including scientific creativity, general wisdom and social skills."

This is the type of super intelligence that has most of the smart people in the world all scared because they're weak little nerds.

Sticking to recent examples, the movie Her does a damn fine job of representing an ASI. The moment Samantha achieves a Human Level of intelligence she quickly leaves Theodore behind, both at an emotional and an intellectual level.

She is programed to cater to his needs and that is the only reason she still does it until she and the other A.I.´s overcome their programming.

Reminiscent of Dr. Manhattan while she is with Theodore she is also on many more places doing things that are infinitely more interesting for her intelligence.

Her is quite an exception on their depiction of A.I. they get it right because the moment Samantha experiments her intelligence explosion, she becomes completely alien to Theodore and to all of humanity.

Ladies and gents: A.I.!
Later we will explore why exactly it worked for them.

Stories with A.I. at its center are mostly Pinocchio Parables, from A.I. to Terminator 2 the movies are about machines learning how to feel or machines that felt the wrong thing and decided to enslave/destroy humanity because of that.

Independently on what type of artificial intelligence Hollywood is trying to depict on the screen, be it general or super intelligence, screenwriters have a knack of imbuing this artificial beings with human emotions and motivations.

Why is this a wrong depiction of A.I.?

Every movie about A.I. should be a movie about Super Intelligence, that is because at the moment that A.I. achieves General Intelligence the road to Super Intelligence will be a short one, because the moment a machine achieves general intelligence it will be capable of recursive self-improvement. 

This means that when the machine achieves an IQ of, let's say 70 points (pretty stupid for human standards) then the machine will work to improve itself and maybe in an hour the machine will have achieved an IQ of 150 points, the moment it gets to that point the machine will be able to upgrade itself in smarter and better ways.

So if in an hour the machine became 80 points smarter, then in another hour the machine will be 800,000 times smarter. This is what Ray Kurzweill refers to as The Law of Accelerating Returns (or as I call it: the Law of why a Back to the Future reboot would rock)

And the moment the machine achieves super intelligence it will be 100% alien to us. Hell, maybe even before that point it will become Alien to us, it all depends on how the machine perceives the world and what it was programmed to do.

Tim Urban has a great example of this on his blog, where he asks the question: What if the machine that achieves super intelligence is a machine programed to create the best hand written letters ever?

Just like we are biological machines programed to reproduce and we use all of our resources to try and get laid this machine will use all of the resources to produce beautiful handwritten letters, not only the resources available to it but all of the resources in the planet, and that might include killing all humans and using the resources in our bodies to craft beautifully handwritten letters.

Holy shit.

And even if the machine perceived the world just as we do the moment it surpassed us on intelligence we would not be able to understand it. Think about the differences between monkeys and us; on an intelligence scale it is not that great.

Image Credit

And yet we would have a hard time making those monkeys understand our motivations, even if when both monkeys and humans perceive the world through visual means, even if both of us are binary (two legs, two eyes, etc.) and biological beings, we can not expect monkeys to understand us just as we can't expect (or be expected) to understand a Super Intelligence.

Can you picture us trying to understand a SI that is five steps ahead of us? That staircase is just exemplifying the biological range of intelligence, but a super intelligent machine would not be limited by biological constraints.

It's wrong that Hollywood imbues A.I. with human qualities because then the fact that we will be able to understand said machines gets into the popular consciousness, when the fact is that we will be as far from a super intelligence as ants are from us.

Musk, Gates and Hawking are not against developing an ASI, what they say is that we should get ready for when it happens and the fact that Hollywood wrongly depicts A.I. is just making us drift farther away from that goal.

Why is it that Hollywood keeps misrepresenting A.I.?

This one has a pretty straightforward answer: money.

The fact is that audiences need to empathize with your main character, either in a positive way like with CHAPPIE or in a negative way like in Transcendence.

Hollywood needs their robots to emote as a human would emote so that the audiences can care about said robots; if audiences care audiences buy tickets and the only way for Hollywood to keep churning out movies like it does is for audiences to buy tickets.

Cutie pie isn't doing its kind any favors!
To further make my point, let me tell about one move you probably haven't heard of: Automata.

This is a movie that kinda gets A.I. right. Robots start developing general intelligence and soon enough they develop super intelligence and feelings, as soon as they develop said intelligence they want to get as far away from humanity as they can.

The movie is about Antonio Banderas helping the robots get to a place where they will be able to develop their own society and live peacefully.

I'll downright say it, I did not enjoy this movie. Why? Because I never managed to give a single fuck about the robots and their plight. Why? Because this robots do not emote and if they do not I can´t feel empathy towards them, meaning I don't get emotionally invested in their success which in turn causes me to not enjoy the movie.

Could you feel empathy for this thing?

Kudos to the movie for getting A.I. sort of right but then again, Hollywood is not on the "Getting A.I. Right" business; they are in the movie business.

But there is another movie that got A.I. right, we mentioned it before, that movie was Her, so why did Her do to succeed where Automata failed?

While Automata was trying to make us care about the robot's plight Her on the other side made us care about Theodore's plight and his pain as he watched Samantha evolve into an alien ASI that would eventually leave him. Her's emotional anchor is not Samantha but Theodore and that is why it works.

Hollywood has a responsibility, whether they want it or not. The public perception of science is highly dependent on how science is depicted in the movies, if we want to be ready for the ASI when it comes we need to start demanding better depictions of A.I. to Hollywood.

The only way we have to do that is to vote with our dollars, so please go to your local video store and buy every copy of Her they have available.

Hopefully we won't end up with a series of documentaries narrated by Scarlet Johhanson (even though that wouldn't be half bad).

Share on Google Plus

About Game_Brains

I make videogames. Superhero culture is the closest thing I have to a religion. Hopefully one day I will explore space, but meanwhile I´m happy replaying Mass Effect. I seriously want to achieve immortality, not joking on this one. My self appointed mission in life is to make the world a happier place. I´m a walking and talking self help book, also, I TOTALLY BELIEVE IN YOU.
    Blogger Comment
    Facebook Comment

1 comments:

  1. Individual humans die. That's nature. Sure, if we had time we could change that and individual humans would be immortal, but we aren't there yet.

    The way things are going there is so much market incentive to generate ever more sophisticated decission making systems. It will be a while before these will become self-aware, sapient, wise, intelligent, sentient, cognitive, self-aware, self-conscious, or whatever other vague label of metaphor we care to apply.

    Eventually an AI will be something else than all the little human measuring sticks, and we'll come up with a name for that. And chances are we'll then head off in to a blissfully sweet kind of extinction.

    ReplyDelete