Thursday, March 26, 2009

exponential microculture

gorknet | Few ideas are so preposterous that no one at all takes them seriously, and this idea - that God, or at least the universe, might be the ultimate large-scale computer - is actually less preposterous than most. The first scientist to consider it, minus the whimsy or irony, was Konrad Zuse, a little-known German who conceived of programmable digital computers 10 years before von Neumann and friends. In 1967, Zuse outlined his idea that the universe ran on a grid of cellular automata, or CA. Simultaneously, Ed Fredkin was considering the same idea. Self-educated, opinionated, and independently wealthy, Fredkin hung around early computer scientists exploring CAs. In the 1960s, he began to wonder if he could use computation as the basis for an understanding of physics.

Fredkin didn't make much headway until 1970, when mathematician John Conway unveiled the Game of Life, a particularly robust version of cellular automata. The Game of Life, as its name suggests, was a simple computational model that mimicked the growth and evolution of living things. Fredkin began to play with other CAs to see if they could mimic physics. You needed very large ones, but they seemed to scale up nicely, so he was soon fantasizing huge - really huge - CAs that would extend to include everything. Maybe the universe itself was nothing but a great CA.

The more Fredkin investigated the metaphor, the more real it looked to him. By the mid-`80s, he was saying things like, "I've come to the conclusion that the most concrete thing in the world is information."

Many of his colleagues felt that if Fredkin had left his observations at the level of metaphor - "the universe behaves as if it was a computer" - he would have been more famous. As it is, Fredkin is not as well known as his colleague Marvin Minsky, who shares some of his views. Fredkin insisted, flouting moderation, that the universe is a large field of cellular automata, not merely like one, and that everything we see and feel is information.

Many others besides Fredkin recognized the beauty of CAs as a model for investigating the real world. One of the early explorers was the prodigy Stephen Wolfram. Wolfram took the lead in systematically investigating possible CA structures in the early 1980s. By programmatically tweaking the rules in tens of thousands of alterations, then running them out and visually inspecting them, he acquired a sense of what was possible. He was able to generate patterns identical to those seen in seashells, animal skins, leaves, and sea creatures. His simple rules could generate a wildly complicated beauty, just as life could. Wolfram was working from the same inspiration that Fredkin did: The universe seems to behave like a vast cellular automaton.

Even the infinitesimally small and nutty realm of the quantum can't escape this sort of binary logic. We describe a quantum-level particle's existence as a continuous field of probabilities, which seems to blur the sharp distinction of is/isn't. Yet this uncertainty resolves as soon as information makes a difference (as in, as soon as it's measured). At that moment, all other possibilities collapse to leave only the single yes/no state. Indeed, the very term "quantum" suggests an indefinite realm constantly resolving into discrete increments, precise yes/no states.