Using a variety of genetic starting points and applying a
range of techniques, both old and new, there is now the potential to shift the balance points and achieve far more of
all goals than is currently the case. But the potential is not infinite and we may still come up against serious constraints
on what can be achieved. Natural selection constantly encounters constraints. For example, the most colourful males may be the most attractive to females but are also most vulnerable to predators (eg Godin & McDonough 2003). The antelope that visits the waterhole and gets eaten by a predator while it is drinking is an example of such a compromise, Natural selection has not evolved the perfect antelope that does not need to drink, can always escape
from predators and lives an infinitely long time. Natural selection, in other words, does not lead to perfection, but to
optimal compromises between conflicting goals, constrained and limited both by underlying genetics and by the real world
(R Dawkins 1982). These constraints include pleiotropy (genes with both desirable and undesirable effects), historical
constraints (such as the blind spot in the vertebrate eye) and the simultaneous evolution of other organisms, such as
predators and diseases, engaged in a constant arms race to run faster or to overcome immune resistance.