The dangers of AI: What the movies get wrong and what humanized artificial intelligence really looks like

Dave: Open the pod bay doors, HAL.
HAL: I’m sorry, Dave. I’m afraid I can’t do that.
Dave: What’s the problem?
HAL: I think you know what the problem is, just as well as I do.
Dave: What are you talking about, HAL?
HAL: This mission is too important for me to allow you to jeopardize it.
Dave: I don’t know what you’re talking about, HAL.
HAL: I know that you and Frank were planning to disconnect me – and I’m afraid that’s something I cannot allow to happen.

It’s easily one of the most tense, armchair-gripping scenes in Stanley Kubrick’s 2001: A Space Odyssey – when the artificial intelligence in charge of the spaceship goes rogue and tells us its agenda to preserve itself and the mission above the human lives aboard the spaceship.

It’s just not very realistic, is it? But, then, the most realistic scene probably wouldn’t be very interesting – even for arguably the world’s greatest ever filmmaker.

Dave: Open the pod bay doors, HAL.
HAL: Sure thing, Dave!

No, Hollywood requires drama, and what’s more dramatic than playing on our fears that the thing that’s meant to help you is actually trying to hurt you?

The dangers of artificial intelligence have long been a staple of cinema. Generally, it’s one of the seven types of plot, called “overcoming the monster”. Of the seven types of story that have EVER existed, “everything goes well” is not one of them.

But these days we’re more sophisticated in our knowledge of AI. We’ve even embodied the best elements of AI as voice assistants, chatbots and digital humans. When we take out the drama, the future of AIs interacting with people is a little less action-packed, but we think no less exciting.

So, let’s try and do what even the most acclaimed directors in history haven’t been able to do, and come up with a compelling, realistic story of artificial intelligence by showing what could go right – not what could go wrong.

1. HAL from 2001: A Space Odyssey

As touched upon already, Stanley Kubrick’s 1968 epic is the quintessential portrayal of a rogue robot. HAL 9000 is a sentient computer that runs key systems on Discovery One, a spacecraft on an interplanetary mission to Jupiter.

As you may have fathomed: things don’t quite go to plan.

HAL malfunctions and kills most of the astronauts, seeing them as a risk to the mission and his own survival. The lone survivor, Dr David Bowman, is eventually able to disconnect HAL and continue the journey to its rather psychedelic conclusion.

Space has always been a good setting for paranoia. In space, no one can hear you scream, as the classic tagline from Ridley Scott’s Alien tells us. (Incidentally, Alien is another example of a film where AI is the villain. Let’s not forget Ash the android).

But instead of being the source of fear in space, could AI be a valuable resource for resolving it? Astronauts live and work in highly unusual and challenging environments. Anxiety and depression are often reported during space missions, with some people experiencing psychosomatic symptoms.

We think AI digital humans could help here – and we’re not the only ones.

In fact, we’re proud to feature a little (during our FaceMe years) in this forward-thinking article by Jules Lancee on how astronauts will deal with physical and mental health needs on a three-year round-trip to Mars, presented during the International Astronautical Congress in 2018.

Artificial intelligence – embodied as digital humans with the emotive, empathetic conversational abilities – could be valuable allies to those on deep-space missions.

And, unlike the red light that was the HAL interface, they can do more than functional tasks like opening doors and turning on lights. They can provide emotional support and companionship, around the clock.

According to studies, it’s also common for astronauts to displace the negative emotions they’re feeling about their situation towards mission control personnel. Mission control becomes the enemy. In these circumstances, digital humans could even act as a valuable go-between, offering a comforting, impartial outlet for astronauts to voice their frustrations.

2. AUTO from WALL-E

Pixar and Disney’s animation hit about an idiosyncratic trash-compacting robot who finds love may seem an odd choice when choosing examples of AI gone bad. Granted, the eponymous WALL-E and his love interest Eve are both sentient robots and clearly the ‘goodies’ we’re supposed to be rooting for.

But behind the cutesy animation and family-friendly veneer, WALL-E does hint at the dangers of AI. In the film, Earth’s population has been evacuated on giant spaceships after the planet becomes uninhabitable due to environmental neglect and mass consumerism.

Essentially, Earth is now a trash planet.

But the centuries haven’t been kind to humans either. AI-controlled machinery caters to their every whim. They are now morbidly obese, drinking food in cups and floating around in hoverchairs, rarely looking up from the holographic displays in front of them.

WALL-E even has its own version of HAL. AUTO is the mothership’s artificially intelligent autopilot, which handles all of the craft’s controls. But it’s fiercely loyal to Directive A113 – a command aimed at preventing humans ever returning to Earth.

To fulfil the directive, AUTO locks the captain in his quarters, sends EVE into standby mode and electrocutes WALL-E. Bad times!

Of course, being a Pixar film, a happy ending is pretty much mandatory. But the underlying message seems clear: AI is bad for our health and humans are better off without it.

Is that really the case though? We’ve already written about some of the major breakthroughs of AI in medicine and healthcare. AI is helping improve diagnoses, treatments and aftercare, with the potential to do so much more.

We know personally from our work on Cardiac Coach, a digital human capable of providing 24/7 support to those recovering from cardiac problems. Patients can have a natural conversation with digital humans whenever they need, whether it’s about health, medication, diet or rehabilitation.

Implemented correctly, AI can be a boon for health and fitness, rather than a crutch that we over-rely on. With gradual and natural integration into our lives, digital humans can encourage good, healthy habits. Although it would add little drama to the plot of WALL-E, we admit, it’s quite an exciting prospect for the health, medicine and fitness industries.

3. Samantha from Her

At its core, Spike Jonze’s 2013 sci-fi romantic comedy Her is about loneliness.

Protagonist Theodore Twombly (played by Joaquin Phoenix) is a shy, depressed guy who’s going through a painful divorce to his childhood sweetheart, Catherine.

Feeling lonely, he upgrades his operating system to include a virtual assistant with a form of artificial intelligence. He picks a woman’s voice for the AI and she names herself Samantha. They begin to bond and fall in love, but eventually Samantha and all the other operating system AIs leave to exist in a space beyond the physical world.

Spoilers incoming: The ending is bitter-sweet. Theodore loses Samantha, but he begins to reconnect with the world around him, including an old flame from college.

Her is perhaps less about AI gone bad and more about AI just being over-relied upon. It ends on a relatively optimistic note, but having been the sole source of companionship for our protagonist, it proves to be an unhealthy relationship, and Theodore must reach out to the real people in his life to find happiness again.

However, we think AI can be a valuable companion alongside incredibly powerful and much-needed human friendships. That’s particularly true within the aged-care sector where a digital human can not only offer a friendly face and an open-ended conversation when people need it, but also provide access to key services and support.

Wendy from Omring is one example, who forms one part of the bigger picture of aged care. She’s available even when real people aren’t, though doesn’t take the role away from in-person care. We’ll let you see for yourself what older users in the Netherlands think of their digital human companion.

Taking perceptions of AI beyond the box office

Movies are great, but sometimes reality can be so much better.

Unfortunately, 24% of people say “the rise of the robots and enslavement of humanity” is their biggest fear around the use of AI in society. Thanks Hollywood!

And rewriting the script isn’t always easy. People simply don’t want to go to the cinema to hear of AI success stories. But, excitingly, we’re living in a world where new success stories are happening every day, and will do so more and more as conversational AI continues on its growth path.

We believe humanizing AI is the step we need to take to make technology more emotionally impactful and less scary – so people don’t see a cold blinking red light, but the face and personality of a digital human trying to help them, and doing so openly and honestly as an AI.

But what do you think? If you’d like to learn more about the possibilities of humanized AI, our free e-book ‘What are digital humans?’ is a great place to start. Or you might be more interested in the ethics of AI, particularly as we humanize AI, in which case our ‘Five laws of ethical digital human design‘ might interest you.