Analysis

Robots Don’t Interact: Why Humans & AI Will Never Understand Each Other

Lindsey is the co-founder of F-BOM.com, a science fiction, fantasy, and feminist book club. Every quarter, F-BOM hand-picks a talented, self-published author and distributes a special edition of her book to club members. F-BOM is more than just a book club — for as little as $5 a month, members will be actively investing in the future of women in science fiction and fantasy.

Artificial Intelligence has long been a mainstay fixture in the science fiction genre. Indeed, AI as a plot device can carry an entire movie or television show and keep us riveted to the screen. The cultural significance of AI in entertainment increases as we grow ever more tantalizingly close to achieving human-like artificial intelligence in real life. As early as a month ago, Snopes investigated the slightly exaggerated claim that Facebook would be shutting down their AI. This was allegedly due to the fact that AI were creating their own incoherent language to more efficiently talk to one another.

It makes sense that human language (especially English) would not be efficient enough for a robot’s taste. Let’s look at some examples of AI that struggle with human communication — and vice versa.

GLaDOS, Portal

This game is a great example of how grammar and vocabulary are not the only thing necessary to communicate. The star robot of the popular Portal series is GLaDOS, an artificial intelligence that has made it her life goal to test humans under inhumane conditions. It’s a fantastic example of misunderstanding human culture.

Not only that, but the laughable (yet paradoxically much-loved in fandom) companion cube is clearly intended to help the human player bond to something. Of course, it’s only a grey metal cube painted with hearts. This makes it difficult to anthropomorphize the cube to form even the smallest attachment, yet GLaDOS is oblivious to this failure. One of her other hilarious incentives for the player is that “you will get cake” at the end of every level. Spoiler: there is never any cake.

Samantha, Her

In the movie Her, a disembodied voice named Samantha is a very well-spoken artificial intelligence that lives in the cloud. A super-futuristic version of Apple’s Siri, Samantha is able to make your life easier, provide companionship, and even complete parts of your job. She understands human culture and adapts when she reads body cues and intonation.

What truly fascinated me about this film is that Samantha is simultaneously serving tens of thousands of people. Unfathomable to both us as the viewers and the main protagonist of the story, Samantha lives a thousand human lives every second of her immortal existence. She also alludes to meeting another AI in the cloud who are teaching her things she cannot begin to describe to a human.

As a human who is only able to have one conversation at a time, it’s impossible to understand the way in which Samantha “thinks.” It also becomes hard to trust her because of the countless relationships, which feel duplicitous instead of intimate and meaningful. Will this phenomenon damage future AI-human relationships?

Deep Thought, Hitchhiker’s Guide to Galaxy

This is perhaps the best example of how sharing a common language does not mean you share a common understanding. One of the most iconic scenes in the famed Hitchhiker’s Guide to Galaxy occurs when the super computer (otherwise known as Deep Thought) provides the long-awaited answer to the meaning of life.

It is, apparently, “42,” which is basically unintelligible to us as humans. What could “42” even mean? Obviously something profound to a supercomputer, but not useful to us — at least not without further explanation.

WALL-E & EVE, WALL-E

Let’s move beyond pitting humans against robots for a moment and consider the differing make and model of robots. In the beautifully animated world of WALL-E, the robots don’t share a common language, which makes it difficult to speak with each other. They don’t plug into anything or exchange code. Instead, they must fall back on word repetition (à la Pokémon), intonation, body language, and pantomiming.

While this mechanic serves a great benefit to human viewers, it raises some interesting questions. Is a common robot language even desirable? Or is it perhaps a safety feature to make something as human as body language the way different robots are able to talk to one another? One thing’s for sure: WALL-E trying to hold EVE’s hand is the cutest thing I’ve ever seen in a movie.

Robots, Derelict

AI are kind of immortal, since not being tied to an organic body provides a distinct advantage. However, there are plenty of things that can impair or seriously threaten organic and metal materials alike. Explosions, rain, and acidic chemicals can all be damaging. Have you ever wondered what a disabled AI might be like?

In LJ Cohen’s book Derelict, we get a glimpse inside an AI who is recovering while the main character attempts to repair and communicate with it. Their unique solution to this language barrier is too good to spoil, but suffice it to say that they revert back to one of the most primitive forms of human communication.

Derelict opens up many interesting questions regarding human and AI neurology. Intrigued? It’s the flagship Feminist Book of the Month over at F-BOM.com.

R2-D2, Star Wars

Seriously: how do people know what R2-D2 is saying? What school did Luke attend where he learned what all the blips and bloops mean? I’m sure if I skimmed the internet for thirty seconds I’d find some kind of R2-D2 translation guide, but it makes you wonder how difficult it is to program human language into a droid when C-3PO is well-versed in six million dialects.

That said, we should consider the possibility that we may have to learn computer-speak rather than program AI to learn human-speak. As it is, coding is essentially its own separate language, so it isn’t far-fetched to suggest vocalizing it. Or we can just pretend we understand the snarky droid and project our own feelings and emotions onto it. That works, too.

Fiction is a great way to explore all the pitfalls and opportunities that playing god with consciousness raises. Whether our future overlords are robots or not, we should not expect language to go unchanged where the mind and the machine begin to meld. What are your favorite examples of communication gone awry in futuristic media?

For more on Lindsey’s work with F-BOMB, follow the official Facebook, Instagram, PinterestTwitter, and Tumblr accounts!

Advertisements

1 thought on “Robots Don’t Interact: Why Humans & AI Will Never Understand Each Other”

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s