In answer to the question "how will we get our news if journalism dies" :
Maybe a site *only* for "scoops", ie. breaking news stories which haven't been seen anywhere else online.
You have some Digg-style voting. Then, after giving readers 24 hours to challenge the *originality* of the story (by finding earlier references), the highest voted (most important, sorta) story of the day wins a cash-prize. If someone finds it elsewhere online, then you don't give the prize. (But you do give something to the guy who found out it wasn't original.)
Basically you do the opposite of Wikipedia's "no original research" policy. You ruthlessly reward "scoopness".
Pay for the prize by charging subscribers a few dollars for a feed which has the stories an hour earlier than they appear on the site. For real news junkies, an hour early may be worth a subscription of 10 dollars a month. (A hundred thousand subscribers gives you a million dollars a month from which you get your prize-money and your costs.)
Sell further analysis services around that feed. Could be automatically generated statistics, could be expert opinion, could be classification, could be ad-supported blog commentary, could be print-on-demand hard copy.
Here's some more brainstorming.
You can offer, each year, grants to those investigative journalists who have scored highly, in the past year.
Prizes could be allocated in different ways ... maybe one big and a lot of small ones. Or a 1st, 2nd, 3rd big prizes. Or you could have two prizes for pure originality / importantness and one for originality / well-writtenness. Whatever virtues you want to encourage in your submissions, offer some feedback and reward for it.
To make it super-easy to contribute, allow anyone to just email or IM a story in.
There might be some issues with, say, a very early but short SMS-type message : "Bomb at King's Cross" vs. a more measured three paragraph mail about the same thing, arriving 5 minutes later. Who deserves the prize? Perhaps segment into different "pools" of different sized text : Micro (the twitter-size), Article, Feature etc. with separate prizes for each. And, sure, that means someone might get inspired by the Micro post to chase and write up an Article-sized one. But that's fine. That's what should be happening.
Similarly you can segment by locality.
What about the big problem of people deliberately feeding *false* information into the system and then it getting voted up (maybe due to political prejudice)? Well, that's not a problem that we have to eliminate altogether. This happens (too much) with the current system we have today. (Ie. newspapers printing false things that suit their biases.) What we can work towards is that a) we obviously won't give a prize to a story which is found to be false, b) those who consistently contribute or vote for false stories can be punished in some way. (Their votes are discounted.)
Update 2 : Of course, the big problem with the above is that you can't simultaneously rely on the mass of reader / voters to debug and add value to your news feed AND sell early access to it. Paying subscribers are least likely to be interested in working on prioritizing and identifying novelty. They're paying for a finished product. So the business model needs to be different.