Michael Weinstock, The Architecture of Emergence.
Emergent behavior is an interesting design concept because it relegates the designer to the role of the unknowing tinkerer. That one cannot possibly conceive of the result of an atomic ruleset multiplied over time means that the ruleset itself is unaware of its affects. Simulation tools begin to enable a looping process where the ruleset is created, the emergent properties are observed, and the ruleset is modified based on those observations. But even here we cannot exactly design results, only tinker until the simulation yields a more desirable affect.
And that's the point. The designer is no longer encoding a specific outcome, but encoding methods of interactions at the smallest scale. Here, to a large extent, the means justifies the end (if there is an "end"). Furthermore, emergence implies evolution and a temporal behavior; inanimate systems designed in this way place themselves in a world of change and time wherein the system may create completely unexpected results, it may become malignant, or it may cease to exist altogether. As J. Huxley describes, in this evolution, "now and again there is a sudden rapid passage to a totally new and more comprehensive type of order or organisation, with quite new emergent properties, and involving quite new methods of further evolution." In other words, when a system is imbued with the capability for change, the patterns of evolution may be haphazard, radical, and stuttering. If this system displays strong emergence, the designer must truly act as a humble organizer of minutia, capable only to simulate the limited possibilities of future states within the scientific void of variable isolation.