TORONTO — In the world of sci-fi movie geekdom, Aug. 29, 1997, was a turning point for humanity: On that day, according to the "Terminator" films, the network of U.S. defense computers known as Skynet became self-aware — and soon launched an all-out genocidal war against Homo sapiens.
Fortunately, that date came and went with no such robo-apocalypse. But the 1990s did bring us the World Wide Web, which is now far larger and more "connected" than any nation's defense network. Could the Internet "wake up"? And if so, what sorts of thoughts would it think? And would it be friend or foe?
Neuroscientist Christof Koch believes we may soon find out — indeed, the complexity of the Web may have already surpassed that of the human brain. In his book "Consciousness: Confessions of a Romantic Reductionist," published earlier this year, he makes a rough calculation: Take the number of computers on the planet — several billion — and multiply by the number of transistors in each machine — hundreds of millions — and you get about a billion billion, written more elegantly as 10 to the 18th. That's a thousand times larger than the number of synapses in the human brain (about 10 to the 15th).
Koch, who taught for more than 25 years at Caltech and is now chief scientific officer at the Allen Institute for Brain Science in Seattle, is known for his work on the "neural correlates" of consciousness — studying the brain to see what's going on when we have specific conscious experiences. Of course, our brains happen to be soft, wet, and made of living tissue, while the Internet is made up of metal chips and wires — but that's no obstacle to consciousness, he says, so long as the level of complexity is great enough. (Most researchers working on artificial intelligence would agree that the "substrate" doesn't matter. That is, it makes no difference what the system is made of. Most philosophers, though not all, would agree.)