In the past, genetic wizardry has been confined to tinkering and tweaking what nature has already produced—taking a gene from a bacterium, say, and inserting it into the chromosomes of corn or pigs. What we're talking about is producing life that is wholly new—not in any way a genetic descendant of the primordial Mother Cell. The initial members of each newly created breed will have no ancestors at all.

And once a biodevice is designed and properly fabricated, the hard work is over—its users can instruct it to make as many copies of itself, by itself, as are needed. Biodevices could churn out any imaginable pharmaceutical drug, including ones that are impossible to produce by traditional chemistry, or are prohibitively expensive today. Similarly, they could create any other kind of chemical or polymer for the production of plastics, real wool or silk—at a fraction of today's costs—or other structural and functional materials that have yet to be conceived. Other biodevices could act as sensitive environmental biosensors, programmed to detect and degrade specific toxic organisms, such as anthrax, or to glow in the proximity of a biological, chemical or radiological weapon.

Venter and Church are eyeing an even bigger prize: a self-sustaining, highly efficient biological organism that converts sunlight directly into clean biofuel, with minimal environmental impact and zero net release of greenhouse gases. What would an ideal biofuel-generation system look like? "The most sustainable source of energy is sunlight and the most convenient products are pipeline-compatible petrochemicals," says Church. "So I would aim for a perennial plant system that secreted pure chemicals—octane, diesel, monomer for plastics, etc.—into pipes without need for further purification."

"They've forged chemicals into synthetic DNA, the DNA into genes, genes into genomes, and built the molecular machinery of completely new organisms in the lab—organisms that are nothing like anything nature has produced."