dextercioby said:
Hunt_mat, that's only half the story. As Hurkyl noted, the complex exponential is multiple-valued, so we must not forget that.
A typo: it's the logarithm that's multiple valued

but you knew that of course.
Anyway, while it is certainly true that a^b has multiple values for complex numbers, mathematicians sometimes pick one value as a principal value. That is, they define
a^b=e^{bLog(a)}
where Log is the principal branch of the logarithm, which is not multivalued (since we restricted it).
We consider the principal values of a^b in the Riemann-zeta function, for example, where
\zeta (z)=\sum{\frac{1}{n^z}}
there we take the exponentiation to be the principal value, and not the multi-valued one. The value
i^i=e^{-\frac{\pi}{2}}
that hunt_mat gave was the principal value. If you type i^i in google, you will see that they also return the principal value.
I'm not saying that anybody did anything wrong here. But I just wanted to tell the OP that there are multiple values of a
b, but that we often restrict these multiple values to get a principal value.