I'm not convinced that the internet has further shortened our attention spans. If I tried hard enough, I might be able to argue the opposite.
I think that the inverted-pyramid writing style for newspaper journalism proves that Nicholas Carr -- whom I respect very much -- might be wrong. We put the salient facts at the top or our stories, because the reader won't make it to the bottom. Many readers can't carry a concept to the end of a prolix sentence.
The internet has not made us more stupid (um...stupider?). We've known for decades that people have short attention spans.
A six-minute YouTube video can contrast the morality of socialism against self-aggrandizement more efficiently than a college professor's dissertation.
Perhaps it's a matter of how convenient is the empathy. Could it be that watching and listening provide a vicarious experience that allows us to fall more easily into understanding? I believe it's natural that we might be regressing to images (and sound) as the primary media. I think cave paintings and oral histories predate the printed word, right? Our brain is already wired for those (mirror neurons, for example).
If attention span is a problem, I believe it's subordinate to competency in media literacy and critical analysis.