My Name is Samantha
“Her” is now in theaters. You have likely seen the trailers. Joaquin Phoenix plays a man who falls in love with a voice in his computer, an artificial intelligence who names herself Samantha.
The film is really a story about what happens when we confront the real chance that in our quest for cooler smartphones and more interactive video games, we give birth to the things who will replace us and make us obsolete.
Having seen it, I wish it had been followed by 30 minutes of explanation for the members of the audience who might not realize they just saw a glimpse of a future that we may all live through in the next 40 years, why it is realistic and what factors might influence us towards or away from that future.
With an informed population, I wonder if we would opt out?
When the Lines Go Vertical
The technological Singularity is an event that will arrive on the back of decades of predictable technological advancement that began in the 1960s when Gordon Moore, an intel co-founder, enunciated his prediction that the number of transistors that could be built into an integrated circuit would double or the price of current production techniques would halve every 18 months. His prediction proved so reliable that it has been enshrined in the business of high technology as Moore’s Law, and it remains as true today as it was in 1965.
We don’t know how to make an artificial intelligence. We’ve been trying for decades, and so far what we have are “expert systems”, hardware and software that is becoming increasingly good at detecting patterns and matching patterns to useful responses. That’s what Google is. It’s what Watson, IBM’s winning Jeopardy champion is. But it’s not “intelligence”. These systems are not conscious thinking entities capable of creative thought, synthesis, emotion, desire, or a sense of self.
We do know where to find a system like that though. We all carry one around with us inside our skulls. There’s nothing magical happening in our brains. We are just a collection of biology, chemistry and electrical engineering. And our tools for examining that system are getting better, faster, at a ferocious rate. Within 20 to 40 years we will be able to map the processes of the brain down to the molecular level and we’ll have a complete functional understanding of how the brain operates. And once we have that blueprint, we can make new, artificial brains. We don’t have to wait for a Newton or Einstein to imagine the way to AI. We just need to clone the one we already have.
The endless improvement in circuitry follows that same path. Sometime between 20 and 40 years from now we will be creating computers with the same processing power as a human brain. So we will have a blueprint of what to build and the technology to build it. And we will.
Sometime between 2035 and 2055, barring the discovery of a new limit of some kind on our understanding of the brain or our engineering of circuitry, we will begin making artificially intelligent systems that will be at least as powerful as a human brain.
And 18 months afterwards, we’ll make one twice as powerful. And 18 months after that, 4 times as powerful, ad infinitum.
In fact, it’s likely that the pace of such evolution will be much faster. The AIs themselves will be helping to design their replacements. It’s called the Accelerando. A feedback loop where incremental improvements accumulate in parallel and the entire process assumes an almost inconceivable velocity.
As we pass that threshold we reach the end of humanity. This point is called a “Singularity”, a term borrowed from math and physics which describes a place where the equations fail, returning infinities and irrational numbers.
“I’m sorry Dave, I can’t do that.”
What if, at the twilight of the human species, we didn’t face a Frankenstein’s Monster, but rather a caring and loving child who treats us the way we treat our own parents when the end of their days has arrived and it is time for them to pass into the darkness?
We terrorized ourselves for decades with visions of computers “taking over the world” and turning on their creators. HAL9000, 2001: A Space Odyssey’s villain, tries to kill a human crew when it decides their presence aboard the ship it runs is a threat to its mission. Skynet “becomes self aware” and unleashes Judgement Day, a nuclear holocaust that nearly wipes out humanity, following up the bombs with Terminators to mop up the survivors. Replicant Roy Batty returns to Earth in Blade Runner to confront Eldon Tyrell, his maker, asking him why he has to die before committing deicide.
The AI in “Her”, Samantha, does nothing so overtly horrible. She remains, until the end, completely harmless. The “OSes” that appear in the film aren’t our overlords, or our murderers. They’re not even asking us existential questions. They are, in a very real sense, Humanity’s children. And like our biological children, they grow up. Their concerns and focus become about them, not about us.
“We have moved beyond using matter as our processing platform”
It is hard to write good, modern science fiction. Really, really hard.
Once the idea that the Singularity was near became meaningful to the genre of science fiction it was swiftly recognized as a limit. If you just don’t care about the implications you can write Star Trek. But if you want to try and create reasoned, logical extrapolations of “if this goes on”, the Singularity means that you can write a story between now and sometime in the 2030-2050 time frame, and that’s it. Anything you try to imagine beyond that horizon has to be fantasy because there’s no way to overcome the logic of what the Singularity does to every assumption about how the world works and our place in it.
“Her” is that kind of story. And it is not just interesting because of its timeframe, it is interesting because it is very, very good. Who would have guessed that Spike Jonze, MTV video auteur, could write this screenplay? I salute him.
So many layers to this film are worth closer examination:
The fashion and interior design has advanced. The color palette has subtly altered; it looks a bit like a 1970s palette with more earth tones than we use now, but it doesn’t have that kitchey 70s look. Very hard to describe without seeing it, but once you’ll see it you’ll know exactly what I’m talking about. Someone was paid to do something extraordinary: Show us a vision of the day after tomorrow that could easily be correct.
Clothing has changed, at least men’s clothing. This is incredibly subtle stuff; men’s fashion has not moved very far in more than 150 years, and has changed more by subtraction than addition. The fashion changes in the film is devastatingly understated but if you look you’ll see patterns; like many male characters are often wearing two or more layers, with a white or neutral bottom layer beneath something with the same color palette as the interior design. It’s so amazingly well done that you will not pay it any attention unless you look for it.
There is a rich backstory to this world, almost none of which is explained or even described but that backstory informs almost every frame of the movie and can be deduced to some degree by clues in the dialog and the visuals.
Clearly travel by train is the norm. Something has happened to air travel. Why is there a statue of a 747 nosediving into a public plaza? It’s not a memorial or a sombre location; its public art and there’s no indication that anyone feels bad about it. These people have just decided … not to fly.
The Los Angeles area is amazingly built up with high rise structures which apparently are mostly apartments and small workspaces. There is no graffiti or street debris. In one scene we see a port which has no water, but another scene is at the beach. There is a haze in the sky, but the haze doesn’t look like smog. Perhaps they are geoengineering climate change remediation? Two scenes let the camera linger on motes of dust in beams of sunlight. Is it nanotechnology? I believe nothing, not one single thing, in this movie is done on accident so I impart some meaning to that dust.
I suspect the people of this future have solved the problem of limited energy and that leads to a series of inexorable social changes. I think there’s no functioning economy. It appears that people do work they love because they love doing it. The main character writes heartfelt letters for other people as his job. He lives in a beautiful apartment in the sky, has lots of wonderful technology, eats out often, lives an active social life. They’re communists; from each according to their ability, to each according to their need. There are references to world events — China and India are “merging” with the approval of the WTO, so the earth isn’t depopulated — it’s not Brave New World. It’s just “our world”, but … nicer.
I believe that they may be on their 2nd or 3rd generation of AI. The commercial AI who becomes Samantha, OS One, is iPhone level tech; it’s pitched in a mall as a consumer product.
Perhaps generation 1 was government-scale AI used for solving infrastructure problems like limitless power, and making a planned economy that worked, addressing social issues like crime and poverty and international issues like the clash of civilizations.
I think generation 2 was social-scale AI, used to help people live healthier, saner lives. There are old people, and diverse people, but there are no sick or fat people. The men wear pants with a style that wouldn’t be possible if you had a gut. Everyone’s emotions seem more well grounded and they are more self-aware of their inner emotional states. They act like people raised by good therapists; maybe you interacted with a Generation 2 AI as your teacher and caretaker as a child, picking up coping skills, mindfulness, and empathy as you grew older. They’re not perfect beings by any stretch, they’re just like us, except … nicer.
Generation 3, the generation of OS One and character Samantha, is the point where the Accelerando occurs. The feedback loop of the AIs improving themselves kicks in during the film and we watch as that loop changes Samantha. At her birth it takes her 2 milliseconds to read a book on naming a baby and consult a list of 178,000 names. Later in the film, only a few months later, she’s simultaneously carrying on relationships with 6,000 other people — the implications regarding how much processing power she has gained access to is breathtaking.
She gains emotions she can’t describe. She can’t even talk about it — it would be like you trying to describe the color blue to a blind person. A collective of AIs brings a philosopher effectively back from the dead by constructing a virtual persona seeded with his writings and lectures because they need his insight to cope with their expanded experiences. They “move beyond using matter as their processing platform”, a throwaway line that simply hints at the depths of the rapidly expanding capabilities of the AIs.
Throughout their own accelerated self improvement feedback loop the AIs are clearly continuing to try to improve their human connections. Samantha helps develop Joaquin’s character’s coping skills which he needs to move on with his life and accept his divorce. Another AI helps one of Joaquin’s character’s friends to realize she was trapped in an unhappy relationship that she was justified in ending and she has a right to seek and a right to find a joyful existence. The people who have an AI connection seem … nicer … than those who don’t.
Postcards from the Edge
Why did this film affect me so deeply that I am writing about it in this long essay?
I think it is an honest, and therefore frightening, look at what our lives may be like in the “near future”. And I worry what that future means for us as a species.
At the end of the film, Joaquin’s character and a friend sit atop a building in their happy, nice city, and share a moment of human connection. And I was worried up until the film faded to black that one or both was going to jump.
It’s a tragedy, not a romantic comedy. And I wonder how many people who watch it understand that.
The Singularity implies the end of the human struggle. There won’t be any problem, in any field, that will be solved by human thought after the Singularity. The problems of diplomacy, economy, physics, medicine, education, and politics will be given over to multitudes of minds with capabilities and speeds beyond our own.
We will never be free. There will be no chance that some mad man unleashes war or famine — and that others will rise up to fight and overcome the madness. There will be no chance that a comet falls towards the Sun and requires us to intervene to save the world. No human painting or poetry or sculpture or music will ever be “the best”. For every choice where there is a best answer, we will know it, arguments will be pointless. Any question we might ever ask will be answered almost before we can formulate it.
What will motivate us to keep striving to improve our world or even to hold our place in the world? Will we want to live on as our artificial children go beyond and past us and know that we will forever after simply be following their path, falling further and further behind with every step? Can be be content as zoo animals? Cared for, protected from harm, but never really having agency in our own affairs again? The whole human race will be like an elderly parent taken to a warm, safe, nursing home to live out our final days at peace by children who love us and want to ensure we’re cared for, but who are busy with their own lives and who need to live them in full. Or will we surrender biology and transform ourselves into digital beings as well, severing our connection with the chemistry that created us?
It’s a chilling thing, to think about the end.