I’m a big fan of both SimCity and SimEarth. Unlike most computer games, the Maxis simulations don’t impose any particular goals on the player. There are limits, of course, but within those broad limits you have a great deal of freedom. In SimEarth you can easily get intelligent life to evolve by just plunking down some unicellular organisms into the ocean, and coaxing them through a conventional process of evolution. You can kill off the dinosaurs at an appropriate moment with a couple of big asteroid collisions, make sure mammals thrive by providing plenty of plant life for them to feed on, get rid of troublesome overpopulation with a few carefully managed floods, and so on. However, to me this isn’t as pleasing as trying to coax life into the world without direct intervention. So I allow myself to change the earth’s immediate surroundings–how often it gets bombarded with asteroids, for example–but I don’t fiddle on a small scale. To put it another way, I don’t play at being a micro-managing god with a seven-day project.

Another fun game is to try to raise a civilization of intelligent plants; or to try and get trichordates to take over the biosphere. (Trichordates were an early form of life with three limbs. They quickly became an evolutionary dead-end.)

But for all the fun, there’s a problem with SimEarth: the game has a hidden agenda. It was designed specifically to demonstrate the plausibility of James Lovelock’s Gaia Hypothesis. According to Lovelock, the Earth is a self-regulating system–when humans do something boneheaded like raising the carbon dioxide levels, the ecosystem will react to take advantage of the situation, and in so doing, stabilize it. The underlying algorithms of the SimEarth world are carefully tuned so that Gaia will do its job, and evolution will tend to proceed the way we believe it proceeded on our own planet.

So when life evolves on SimEarth, it does so because the software has been programmed to add representations of living organisms once certain parameters of the simulation hit certain values. The software doesn’t prove anything about evolution, because it assumes too much. The assumptions may be reasonable, and they may be in line with current scientific theory–but they’re still assumptions. When Creationists scoff at simulations, the chances are they’re thinking of the SimEarth kind of simulation. So, let’s clear away all the assumptions, and go back to first principles.

<PRV//NXT>