As Jack D'Aurizio wrote, you have to be specific about what kind of infinity you're talking about. The easiest to imagine is probably countable infinity. For a finite vector you are probably used to writing $x=(x_1,x_2,x_3)$. You can also imagine this as a function, though:
$$x:\{1,2,3\}\to\mathbb R\qquad i\mapsto x_i$$
So in a sense, $x$ is a function from a finite set of indices to the set of real numbers, returning the coordinate for each index. This generalizes easily to $\mathbb R^{\mathbb N}$:
$$x:\mathbb N\to\mathbb R\qquad i\mapsto x_i$$
Now you have a countably infinite set of indices, and for each index you have a real coordinate. But while you are at it, if you think of a vector more like a function, then you can just as well write
$$x:\mathbb R\to\mathbb R\qquad i\mapsto x(i)$$
and suddenly you have $\mathbb R^{\mathbb R}$, a vector whose dimension is the cardinality of the reals. Treating functions as vectors is probably the most important application of infinite-dimensional vector spaces.
So as you asked about geometric interpretation, I'd think about functions with a geometric aspect to them. One typical example of an infinite-dimensional vector space would be the sine and cosine waves you use as base functions for a Fourier transformation. Then each Fourier coefficient is a essentially coordinate in the corresponding vector, with these waves as the basis. Perhaps others can come up with even more geometric applications.