Last post, someone asked whether we've gone beyond molecules to engineer development of an organism. Molecules are what runs development--molecules made by genes. Consider the homeobox HOX genes (which Charlie mentions in Glasshouse). The same fundamental set of HOX molecules turn on development for the fruit fly, the mouse, and the human.
Picking a drug-producing bacterium off a bar counter is one thing. But would we pick a baby because it makes one molecule?
We already do. Suppose your child has an incurable defect, a single broken gene that fails to make one protein. In some cases, the child can be "cured" by stem cells from the umbilical cord of a matched relative. To get the stem cells, you fertilize in vitro, and grow the embryos in a dish. Then pick the one with the right gene that makes the right protein. So that embryo gets implanted, and when s/he is born, the umbilical cord blood cures the sibling. But what about those other embryos in the dish?
In The Highest Frontier, parents pick genes to make their kids look like Paul Newman; and so the entering class is full of Newmans. But to make synthetic babies, why stop with what's human? I once asked Francis Collins, before an audience of a thousand industrial chemists, whether we should improve humans using chimp genes, since the chimp has a stronger immune system. He gave me a look of horror, then took the next question.
Whatever genes we choose, cord blood won't always do. Suppose it's a brain defect, or a cancer? Then we need to carry the gene in with something really good at infecting cells--HIV. The main media refused to actually name AIDS virus for several weeks; it took a comic strip and SF blogs to say it. Of course, the lentiviral vector can't be one that actually causes AIDS; it has to be stripped down and modified for safety. The way you make it safe is (1) gut out the AIDS-causing genes, (2) add animal virus regulators, such as from woodchuck hepatitis virus and cow respiratory virus. Just hope you're not a woodchuck.
Would you like to have chimp genes to fight cancer and malaria? Or maybe kudzu genes for antioxidants?