Artificial intelligence is the Holy Grail of secular culture, and the voice we have given AI is the voice of the HAL 9000 computer, the lovable, murderous central character of Stanley Kubrick's 1968 epic, 2001: A Space Odyssey.
To many of us, HAL is the voice not only of technology, but of deity, too. He has informed and inflamed our imaginations, from Palo Alto, California, to Cambridge, Massachusetts, and from Austin, Texas, to Tsukuba, Japan. In fact, for millions of technophiles, HAL has become a sort of standardized benchmark denoting the difference between what we hope for from technology, and what it is we might actually receive.
In the recently revised Hal's Legacy: 2001's Computer as Dream and Reality, (MIT Press, $17.50), editor David Stork has picked up on this theme with gratifying lucidity. He presents 16 essays that illuminate the terrain lying between science fantasy and technological fact: Was HAL right? How realistic was HAL? And how close to the mark did Kubrick and collaborator Arthur C. Clarke come when they forecasted a supercomputer that reads lips, runs a spaceship, has reasoning, emotions, ethics, and enough intelligence for seamless man-machine interaction?
As it turns out, in many areas they were dead on or far ahead, and in others, way off the mark. But the fact that we even care so much speaks volumes about 2001's artistic success, as well as about the boundless anima with which we invest the machines we build and use. And if you've read even this far, I suspect that you've not only seen 2001, but have also, like me, seen it many many times (about 18 to 20, I figure.)
Beside the kapow visual effects in the film, what about more subtle layers of scientific advance exemplified by the entity of HAL -- AI, interface design, speech recognition, and computer chess playing? Or fuzzier areas such as computer ethics and emotion? Thirty years later, does HAL leave us annoyed, as though we've just walked out of Starship Troopers at the local twelveplex? No.
A more modern HAL, made in 1998, might have tiny digital cameras for eyes instead of the cycloptic red crystal. He would probably be accessed with infrared equipped palmtop and laptop devices rather than through clunky control consoles. (Clarke and Kubrick weren't quite prescient enough to foresee the Macintosh interface or the PC revolution.) A modern HAL might have speech recognition and speech capability, but it is doubtful HAL would have facial recognition skills, let alone lip-reading ability.
Could a modern HAL beat Dave at a game of chess? Probably. Would HAL be afraid of his own death? Well -- that takes us to the whole question of "What is the current state of AI?" The answer, according to this book, is that we have a long, long way to go, and of all the fields in computer research, this is the boggiest going.
Some of Legacy's pieces glide along seamlessly like an information-dense Vanity Fair Q&A with, say, Courtney Love -- as in an interview with Marvin Minsky about his role as scientific adviser on the set of 2001.
Other chapters are crabby and dense with information, but no less interesting, as in Doug Lenat's essay on whether HAL displayed common sense: "HAL had a veneer of intelligence, but in the end he was lacking in values and in common sense, which resulted in the needless death of almost the entire crew." Grumpikins!
In the end, HAL's Legacy offers readers a surprisingly comprehensive overview of the state of the state of computer development, taking readers to a heckuva lot of places they would otherwise never go. It tells us we have a strong impulse to not only Pygmalionize our creations into life, but that this urge to anthropomorphize and deify both enhances and hinders the development our efforts into future generations of machine intelligence. Is the most we can expect out of AI a smarter, faster version of ourselves? Or will we manufacture something transcendent?
That's the big question. In the meantime, if film directors would be even a fraction as considerate as were Kubrick and Clarke in the making of new films, the local Cineplex would be a much happier and inspiring place to visit.
Douglas Coupland's most recent novel is Girlfriend in a Coma.
