Speaking as a citizen, the problem with the US's newest brand of digital weapons, is that they can be used on US population under the radar and w/o killing anyone. This may justify legally their extended use for surveillance unfortunately. The development of the atom bomb and chemical weapons had no "convenient" use on the USs own citizens, and they clearly couldn't get away with it. However, these weapons do, and they are being developed with all the same force, purpose, and financial backing as the a-bomb and chemical weapons were ~100 years ago.
A couple of ideas.
On revenue:
I think dredmorbius really covered it well. And I want to emphasize the infrastructure part. What about selling hardware that focused on the receiving endpoint? The target demographics probably have literacy/language issues with the content, so providing dedicated hardware to read text (instead of broadcasting actual audio, saving bandwidth), translate it, and do text-to-speech could be an example of some additional products to sell. Sell an app version too for middle-wealth countries. Selling infrastructure items doesn't create that conflict-of-interest in my mind that paying for content does.
Jumping off of the classifieds idea. What if you charged a flat fee, the same flat fee, for every submission regardless of what it was? If submitters truly have altruistic intentions, then this small flat fee is a small hurdle for the little guys, and probably a dead end for the very wealthy to try and overpopulate the system with whatever they want. So that the pressure is not on you to arbitrarily regulate the income sources. Decouple money from content is my main theme here.
On content:
I've never liked one-dimensional voting systems that are simply popularity contests. It's not that hard for a organized group to up-vote their propaganda (one group of many) while the altruistic content submitters are all left by the wayside (many groups of one). What I'd like to see is voting be not for submission, but for classification of content. It could be instructional, news, public service, propaganda, entertainment, etc. Then let your paid submitters be the ones to vote separately on ratios of these categories of content that finally get broadcasted. And only content that has been reviewed and voted for past a certain threshold gets proposed for submission. Right now, looking at the stream, I am already biased AGAINST stuff that is paid. Disclosure of that fact is great obviously, but what if a private org really is producing great content? I'll probably never check it out. A community filtering/classification of the content would put my mind at ease as the same standards of review would apply for paid and free content.
They've "baked in" a corrupting factor to the content: money. IMHO this is what soils services like Facebook, Twitter, Angies list, etc. Places like archive.org and wikipedia are flat donations (I believe).
Subtract 6 years, and I'm in almost your exact shoes. I'm a professionally trained musician and music teacher, and while I love those activities and would never want to stop doing them, I always wanted to expand my abilities to include software design as well (plus, regular and/or better pay is always a wonderful perk). So I'm in the middle of doing what so many on here already suggested:
As far as the learning is concerned, constantly switching between different sources above to hear the same information from different resources and taught in different styles is crucial I think. I know this from my formal teaching background and feel comfortable with the learning part of this entire venture.
The second part of your questions though about finding work I'm not so sure about as I'm not there yet. I personally have chosen python as my language to learn because I've found that:
1)It makes sense to me after trying out a couple of different languages on codeacademy
2)The clean syntax and forced indenting structure allows me to focus more on the problem domain then the language itself (or which indenting/bracket style to use). This can be said of any language I'm sure as people have different preferences and perspectives, but this was clearly evident to me early on.
3)With the discovery of Django and so many other python based tools, I realized that I could actually achieve a lot by specializing in python. But of course I need to include html, css, and perhaps random java script as well to get a dynamic web page going, but at least Django frees me from needing to get into php.
4)I personally have a creative mind and less of an engineering mind. Learning C or some other lower level language would drive me nuts I think AS A FIRST LANGUAGE. I'm able to casually learn about different algorithms and data structures in python without having to fully master them just to make a basic program (as I would in C for example).
Now my goals might be different from you, I have a long list of projects and ideas that I would like to do on my own that could possibly make money, so I'm making all my decisions from a passion perspective, not as a supply/demand one. But maybe that's not quite a bad idea after all? You're already starting from scratch, why not try your own projects first doing them in exactly the languages you want to do them in? I've always felt that there are always plenty of people doing the "standard learning" track for the "standard jobs" that the market demands. If your trying to follow a market, you'll always be at the whim of the market. Why not make your own market?
I always come back to this quote:
“Don’t ask what the world needs. Ask what makes you come alive, and go do it. Because what the world needs is people who have come alive.”