Film & Screens » Cinema

Bodies electric

What is AI?, I query.

Artificial intelligence is a branch of engineering and science devoted to constructing machines that think, Chatbot replies, seemingly educated on the subject.

More questions: Does artificial intelligence exist?

It does now :)

Chatbot bothers to insert a smiley face (colon + right parenthesis) at the end of the sentence. Intriguing.

So, is the movie A.I. Artificial Intelligence realistic?

It seems we have something in common, Chatbot retorts.

Huh? Chatbot, the tour guide for the official movie Web site (www.aimovie.com), is confused. Or maybe its response symbolizes the parallel between synthetic life on the silver screen vs. 21st century incarnations.

Today, a toaster borders on complete meltdown during each operation; refrigerators are merely devices to chill perishables and Mountain Dew; and your surround-sound rig, elaborate and expensive, merely boasts audio. All electronic devices, being kitchen appliances or 1,000-watt amplifiers, lead a single-task existence: A blender blends; a dryer dries.

Tomorrow, however, will explode electronics into multicelled, thinking machines. In 20 years, your toaster could grow a brain. Imagine if it could sense lightning storms, unplugging itself from the wall to avoid a fatal shock. It could even memorize your favorite grade of toast and then digest the burnt, crumby residue after each use — the self-cleaning gadget of a new dimension.

Your refrigerator could warn you when a certain carton of milk is going to expire. It could go shopping automatically, ordering food online. A near-empty bottle of ketchup, aged and crusty, could trigger the delivery of a fresh container of Heinz. Even your stereo system could warn listeners of decibel levels and noise-pollution rates. (“Excuse me, Jon, but if you continue to listen to music at this volume, you are at risk for hearing loss.”)

Lover boy

David, a curious boy programmed as an 11-year-old, is birthed from a diagram. A team of dedicated cybertronics technicians constructs him from miles of wire and sensors. They pad him with manufactured flesh. They paint on birthmarks. They score his voice.

This idea, the opening situation of Steven Spielberg’s A.I., grows into something of grandeur. David is not just another household maid or sex-toned pleasure bot. He is a boy with one purpose, natural by all intentions — he has been built to love. A complex experiment in human compassion, David’s sole purpose is to bond on an emotional level with a mortal mother, trapped in sorrow. Yet he was not born of her womb. He is an outsider on many accounts.

Thus, the question is posed: Can an artificial being that looks, sounds and feels like a human be accepted as organic? David, for example, is unaware of the basic functions of humanity. He cannot digest food, therefore he pretends to eat. He can’t eliminate waste from his body. He doesn’t possess wit or exercise spontaneity. Butterflies don’t flutter in his stomach. Cancer doesn’t invade his brain. His memories are permanent — they do not fade.

Flaws and ticks

“The more one pushes against AI’s internal logic, the more one appreciates the profound difference between human beings and any sort of computers that we can even imagine,” Phil Agre remarked awhile back in an e-mail issue of his Red Rock Eater Digest. His views are not necessarily authoritative — he is an associate professor of information studies at the University of California, Los Angeles, with an AI background — but they are well-justified.

Agre argues that artificial intelligence is falsely advertised: “One of the really profound, long-standing problems with AI is the conflation between the mind and the world.”

Present-day AI, along with the more technologically advanced — but ultimately flawed — AI presented in Spielberg’s film, is built on a foundation of routine. Redundancy is key in developing an artificial life. Basic personality clicks, like the stutter of a voice or a wheezing breath, are organic in nature. If translated into artificial life — into a microchip — something as simple as a stutter would lose its randomness. In a robot, any similar mannerism could not be influenced by the outside world. A robot installed with a drooling deficiency, for example, could never be cured through conditioning therapy. It would drool incessantly, possibly set in rotation by a cycle (to emphasize: drooling every 15 minutes, on the most basic level). Once programmed, drooling or stuttering would be routine.

In other words, the human mind is not a hard disk. Tasks such as driving a car or brushing your teeth cannot be broken down into a sequence of tiny events. Humans are bound by a moment-to-moment existence.

In David’s case, he is programmed to love a specific person. In reality, humans fall in love. There is no step-by-step pamphlet — there is no process to love.

Flesh books

“Putting functional reasons aside, there is another motivation for building robotic creatures — to better understand ourselves,” Cynthia Breazeal, postdoctoral associate at the MIT Artificial Intelligence Lab, writes in her essay, “In Our Image.” Agre also touches lightly on the same idea, but from a more extreme angle. Based on the concept of moment-to-moment living, we can conclude that artificial intelligence will probably never achieve mechanical symmetry with human beings. So why bother constructing AI modeled after humans at all — except for slave labor, of course?

Breazeal expounds: “Robots serve as a mirror, reflecting our humanity back at us as we interact with them and they engage us. As we look at these mirrors, we can better see ourselves — scientifically, socially and ethically.”

In retrospect, it’s more feasible and much more convenient to dissect flesh-and-bone specimens — a guideline that doctors have employed since the invention of the scalpel. Studying a mechanical composite may offer a varied perspective, but in the history of science, it has always been common practice to inspect subjects at point-blank range — like the contrast between admiring a photograph of Mars and actually landing on the red planet.

Watching a child mature from birth to adulthood seems like a more informative and productive study than the alternative. Observing a robot boy like David would surely be a pioneering investment, both profitable and worthwhile, but only if the motives are solely robotic. The human brain is not a hard drive. To learn about the maturing of a child, studying an actual child is a necessity.

Accessory blender

“Some day our synthetic progeny might cross the threshold from the inanimate to the ‘living,’ from being an automaton to becoming a sentient being,” Breazeal continues, suggesting a more elaborate direction for the development of AI.

Perhaps it’s not an utterly independent robot being that will emerge from future developments. In light of a major mortal flaw — the decay of organic material such as bone and tissue — AI may transition into the realm of the cyborg, the definitive life-cocktail of natural engineering and mechanically enhanced counterparts.

Eventually, humanity could cross the boundary into immortality if technology sufficed. Organs could be replaced with ease. A man could plug in a dozen synthetic hearts before reaching the end of the line. Lung cancer? The underlying concepts of AI may be the premier warrior against disease in the coming centuries — maybe even sooner, considering the rapid evolution of knowledge.

Until then, the magic of cinema is the only gateway we have to an outlook on artificial intelligence. And time the only fortuneteller.

Jon M. Gibson is currently constructing an AI model of himself. Within a few months, it will be writing articles for him. E-mail him (the robot) at letters@metrotimes.com

comment