As you can tell from my handle, I'm an economist (although I no longer live in Virginia). As such, I see the world through the lens of social science. One of the great lessons from this election is that social scientists (and those like Nate Silver who use the tools of social science) are better at journalism than journalists. For journalism to be useful to the public, journalists need to adopt the tools of social science in their work. Follow me below the orange symbol for the details.
The core problem with modern journalism is that journalism is still designed for a world in which information was scarce. Before the widespread use of the Internet, it was hard for the average citizen to gain information about issues that affected their lives. Journalists were useful because they disseminated basic information for the public. They interviewed public officials and other experts, and printed this information for the public to read. This was the only way most people could get news, so journalists were important. Good journalists also did investigative journalism, where they found information that wasn't publicly available and informed the public about problems or scandals. This was incredibly important, and made journalists incredibly important for the maintenance of a healthy democracy.
With the advent of the Internet, information is no longer scarce. I can use Google (or any other decent search engine), and find most information I'll ever need. If I can't get the information I need on Google, I can call up a university professor or other researcher and they can help me get the information I need. So, journalists are no longer needed to get information to the public. Instead, journalists are needed to process the enormous amount of publicly useful information in a way that is useful to the public. Journalists fail miserably at analysis of publicly available data.
This election was a key example of how social scientists are better at journalism than journalists. Many journalists/pundits devoted months to figuring out who was going to win the election. They failed; on Election Day, most pundits thought the race was "too close to call." However, poll aggregators like Nate Silver (538), Sam Wang (Princeton Election Consortium), and Drew Linzer (Votamatic) wrote that Obama was going to win a clear victory with 303-332 electoral votes, which he did.
For modern news organizations to be useful to the public, they are going to have to adopt the tools of social science. They will not be able to do this with many of the journalists they currently employ. For example, IMHO, the two most interesting pundits the New York Times employs are Nate Silver and Paul Krugman. Neither are journalists - they just write for the NY Times! In contrast, pundits such as David Brooks and Thomas Friedman provide little value to their readers.
News organizations still need journalists to go to places where the average citizen can't go, and to follow up on information that the average citizen can't follow up on. But we need journalists to acknowledge the publicly available information that exists and use that information, rather than assuming the "inside" information they have from that access they have to public officials in always better than information that is publicly available. To do this, news organizations will have to use fewer full-time generalists and more part-time specialists. And when publicly available information provides difficult questions, they need to both get analysis from experts and provide experts with their own platforms to explain our increasingly complicated world.
Journalism can survive and thrive in our rapidly changing world. But to do so, it needs to provide analysis we can use, rather than poorly formed opinions devoid of any real expertise.