Spielberg catches Kubrick's Baton: A Review of
The androids and other intelligent machines in "A.I." represent well-grounded science futurism, says AI pioneer Raymond Kurzweil.
Originally published June 18, 2001 on KurzweilAI.net.
Stanley Kubrick developed his ideas for a movie to be called "A.I." for over ten years, passing the baton to Steven Spielberg upon his untimely death. As was his working style, Kubrick did not write a screenplay, but kept copious notebooks of ideas. The task of carrying Kubrick's conception to fruition presented Spielberg with a singular opportunity, but also unique challenges, the most obvious being how to meld Kubrick's dark visions with his own affirming perspective.
Both filmmakers capture the intricate dualities of life, but Kubrick tends to wallow in the enigmatic crevices of humankind's ardors, from Strangelovian underground sanctuaries to, well, cybernetic psychosis. While Kubrick is likely to reveal the madness that lurks beneath a façade of normalcy, Spielberg's capacity is to show us the humanity that survives human madness. Even in Schindler's List, we are continually able to gain a measure of comfort from the passionate portrayal of one man's exercise of heroism that carries us through to the last moving scene.
So one question I had going into the advance screening of "A.I." was whether we would hear both voices, whether this remarkable and unusual collaboration would preserve the seemingly disparate outlooks of two legendary artists.
My other salient question was whether this portrayal of a world of "strong" AI would reflect what I would regard as well-grounded science futurism, or if it would devolve to the usual facile dystopianism or sentimental utopianism.
I am pleased to say that the movie succeeds in both of these key dimensions. Spielberg fans will certainly be pleased with the beautiful cinematography, imaginative effects, and the continual surprises of the story line. We see allusions to many former Spielberg movies from the story of the intelligent but not altogether human hero's desire to call (and to go) home to the "Close Encounters"-like portrayal of the far-into-the-future nonbiological intelligences.
But putting on my hat as Kubrick fan, I was also quite satisfied as we encounter Clockwork Orange-like dreamscapes and Kubrick's vintage brooding colors. As in 2001, the most human characters are the machines, with the sole exception of David's "adopted" mother.
In A.I., the AI's are neither evil nor particularly destructive. Indeed our sympathies are usually with them, at least mine were. It's the humans who express the base emotions of destructive jealousy as they taunt and ultimately destroy the stray androids in the Coliseum-like spectacle of the "Flesh Fair." Here Spielberg has the opportunity to present the key issues of the movie and, indeed, of the 21st century. Are the independent robots who rummage through junk yards searching for usable spare parts to enhance their lot a threat to humanity? Are they becoming "too smart, too many, too fast," as their human tormentors claim?
Spielberg makes the point that it's not the machine-like quality of the machines that is threatening, as we have become quite comfortable with machine-like machines. It's the potential for amplifying our human nature that is the most menacing, at least those aspects of human nature displayed by many of the humans in the film: the cruelty of the purveyor of the Flesh Fair, the betrayal by the human who frames the Gigolo Joe android for murder, and the greed of the William Hurt character who creates David but attempts to rob him of his individuality.
The Flesh Fair scene also presents the important issue of cruelty to machines. In my lectures I often point out that we don't worry much today about causing pain and suffering to our computer programs, a line which is usually good for a few laughs (at least for an after-dinner audience). Spielberg applies the power of the cinema to show us just how compelling an issue this will actually become.
We meet three generations of AI's in the movie, with the last generation far in the future. Even the first generation we meet is sufficiently human, in the positive sense of the term, to fear their capture and to help protect each other from the angry prejudices of threatened and threatening "real" humans.
David is presented as a more advanced android, one that has the capacity to learn to love. David will become a "real boy" when he becomes capable of this higher emotion. I'd have to say that this matches my own perspective, that we will come to accept nonbiological entities as "human" when they are capable of understanding and expressing our most subtle emotions.
What do emotions have to do with intelligence? In my view, our emotional capacity represents the most intelligent thing we do. It's the cutting edge of human intelligence, and as the film portrays, it will be the last exclusive province of biological humanity, one that machines will ultimately master as well. By the way, if David wishing to become "a real boy" sounds like a familiar fairy tale, the movie makes the allusion and metaphor of Pinocchio explicit. Even early in his development, David is sufficiently appealing that he wins the sympathies of the Flesh Fair spectators, much to the dismay of the master of ceremonies, who implores the audience to "not be fooled by the talent of this artistry."
In the third conception of machines that the movie presents, we see entities that are supremely sublime. I've always maintained that we will ultimately change our notion of what it is to be a machine. We now regard a machine as something profoundly inferior to a human. But that's only because all the machines we've encountered are still a million times simpler than ourselves. But that gap is shrinking at an exponential rate, and the movie examines what I believe will be the last frontier: mastering our most noble emotions, a capability displayed by only one human in the movie and sought by at least one machine. I won't give away the movie's ending by revealing whether David is successful in his quest, but I will say that at one point he does display a decidedly inhuman degree of patience.
It was also my feeling that the very advanced entities we meet later in the movie are displaying a noble character that is life-affirming in the Spielbergian sense. I have also maintained that future AI's will appreciate that they are derivative of the human-machine civilization, and will thereby revere their biological ancestors. This view is supported in Spielberg's conception of the most advanced machines that we meet in the film.
David seeks to learn to love in order to be acceptable to his mother. But the film is neither maudlin nor predictable, since it takes many unexpected turns not only in the plot line but in our understanding of the parameters of David's world. This allows Spielberg to display his playful imagination, which I'm sure kept many graphics scientists busy creating the inventive effects. In the movie's sole exploration of virtual reality, we meet Dr. Know, a hyperactive Einstein-like holographic projection who seeks to guide David and his intelligent stuffed bear, both of whom have since teamed up with Gigolo Joe.
At first, Dr. Know seems too literal, as if he were a contemporary search engine, but we ultimately discover that Dr. Know does understand David's plight and gives him some critical hints. The interaction with Dr. Know also illustrates another keen insight about the future, that the only thing of value is information and knowledge.
The movie accurately portrays erotic applications as at least one vital driving force pushing the technology forward. As Gigolo Joe points out to one of his clients, "Once you've had sex with Mecca, you'll never want a 'real' man again." Mecca, as in "mechanical," is the movie's clever designation for machines. One gets the sense that it was the machines in the movie who came up with this label.
The language used and the depiction of the sexual situations should be comfortable for parents and their young teenage children, in keeping with the PG-13 rating that a film such as this must have. Discussions on Internet lists indicate there was some struggle with the film rating board about the rating. Could it be that like the anxious humans in the movie, the members of the rating board find the very idea of sex with machines to be unduly provocative?
Is this science futurism or idle science fantasy? I believe that the androids and other intelligent machines that we meet in the movie, although imaginative, do represent well-grounded science futurism. We will indeed meet nonbiological entities with the range of intelligence that we encounter in "A.I." We already have hundreds of contemporary examples of "narrow" AI, that is, machines that can perform well defined tasks that we regard as examples of intelligent behavior when performed by humans, ranging from diagnosing blood cells and electrocardiograms, guiding cruise missiles, solving mathematical theorems, playing master-level games, making financial investment decisions, and many others.
It is true that machines today do not yet have the subtlety or range of intelligence that humans display. However, within thirty years, we will have completed the scanning and reverse engineering of the human brain and will be able to instantiate the templates of intelligence that we discover through this endeavor in nonbiological thinking substrates. Some of the "strong" AI's that result will be manifest in human-appearing robots such as those we meet in the movie, while others will take other forms.
One could certainly nit-pick anyone's detailed imagining of the future. For example, in my conception, we'll have images written directly to our retinas so we won't need flat panel displays, let alone paper books; roads won't need visible markers; and circuits won't be etched on printed circuit boards. The future will, in my view, include more virtual reality and more embedded intelligence in everyday objects.
But on the important questions, I feel that Spielberg, armed with Kubrick's decade of notes, got it right, particularly in the idea of the mastery of our higher order emotions as the defining challenge of AI, and the ultimate definition of humanness.
| 
