I have argued that the similarities between living systems and actual computational systems are too overwhelming to dismiss. I've suggested that many of the differences between manufactured computation and natural living systems, both superficial and substantive, have arisen from the complementary circumstances surrounding the origins of the two technologies, but that both approaches must address the same imperatives and are therefore on converging evolutionary paths.
If these arguments even mostly hold up, then we can predict major changes in future architectures of manufactured computation. A recurring theme here has been that many of the defining claims for digital computation and communication--ranging from `instant communication' to `frictionless commerce' to `location transparency', and possibly even the notion of `general-purpose computing'--simply are too good to be true, having been purchased at the expense of utterly ignoring the basic tenets of self-versus-other and local self-reliance. It will not continue this way. Even though many parties would like to have control over individual hardware systems--ranging from software and hardware manufacturers to internet and application service providers to governments and regulatory agencies--in the end the geometry of physical space will assert itself over `cyberspace' as computing systems become aware of themselves and their universe.