Let's say i have a linear transformation $T: V \longrightarrow V$ and a polynomial $g$.
I know that every eigenvector of $T$ with eigenvalue $\alpha$ is an eigenvector of $g(T)$ with eigenvalue $g(\alpha)$. Does it necceserily mean that every eigenvector of $g(T)$ is also an eigenvector of $T$?
Suppose that $T(x,y)=(x,-y)$ and that $g(x)=x^2$. Then $g(T)(x,y)=(x,y)$ and then every non-null vector of $k^2$ (where $k$ is the field that you are working with) is an eigenvector of $g(T)$. But it is not true that every non-null vector of $k^2$ is an eigenvector of $T$.